Title of invention: Elevator analysis system and elevator analysis method
Import by reference
[0001]
　This application claims the priority of Japanese Patent Application No. 2017-209721, which was filed on October 30, 2017, and is incorporated into the present application by referring to the contents thereof.
Technical field
[0002]
　The present invention relates to a technique for analyzing a group control elevator.
Background technology
[0003]
　In a relatively large building, multiple elevators are installed side by side to improve the transportation capacity of elevators, and at the time of call registration at the landing, a system that selects and operates the most suitable basket is introduced. Furthermore, as the size of the building increases, the number of elevators that are installed side by side also increases, and the group management device appropriately controls these elevators to improve services such as waiting time for users. At this time, the group management device attempts optimal control by predicting the elevator usage status using the operation data and the like.
[0004]
　In Patent Document 1, the usage demand is predicted and analyzed from the number of passengers in the car by using the feature amount of the upward boarding ratio, the feature amount of the upward getting-off vehicle, the feature amount of the downward boarding rate, and the feature amount of the downward vehicle getting-off.
[0005]
　In Patent Document 2, a camera is installed in the elevator hall to count the number of people in the hall. When predicting the number of people waiting at a certain time, the average value of waiting times at the same time during a certain past period is used as the predicted value.
[0006]
　In Patent Document 3, the number of passengers in the car is used. The current congestion status of the building and past usage history are used to predict the congestion status.
[0007]
　In Patent Document 4, a camera is installed so as to image toward the entrance of the building from the front of the elevator hall, and when a person approaching the elevator hall is detected, a car is dispatched.
Prior art documents
Patent literature
[0008]
　　Patent Document 1: Japanese Patent Laid-Open No. 2014-172718
　　Patent Document 2: Japanese Patent Laid-Open No. 2015-9909
　　Patent Document 3: International Publication No. 2017/006379
　　Patent Document 4: Japanese Patent Laid-Open No. 2000-26034
Summary of the invention
Problems to be Solved by the Invention
[0009]
　Group management elevator control is limited because it is not possible to know the visit time, boarding floor, alighting floor, and the number of people who will be using the elevator at the landing.
[0010]
　In Patent Document 1 and Patent Document 3, since the number of passengers in the elevator car is used, the status of the boarding floor and the exiting floor are known, but the status of the hall is unknown. Therefore, it is difficult to control according to changes in the landing.
[0011]
　Further, in Patent Document 2 and Patent Document 4, since the camera is installed in the hall, the situation of the hall can be known, but the situation of the boarding floor and the getting-off floor is unknown. Therefore, it is difficult to control according to the conditions of the boarding floor and the getting-off floor. In addition, the cost of installing the camera is additionally generated.
[0012]
　The purpose of the present invention is to obtain the visit time, the passenger floor, the passenger floor, and the number of passengers who will be using the elevator from the data and who will be using the elevator. It is to implement control to reduce dissatisfaction from the user such as waiting time.
Means for solving the problem
[0013]
　In order to solve at least one of the above problems, the present invention is an elevator analysis system including a processor and a storage device connected to the processor, the storage device being an object of control. The number of persons who appear to use the elevator at the landing of each floor of the group is kept, and the processor predicts a future number of persons from the number of persons held in the storage device. The operation rule applied to control the operation of each of the baskets belonging to the elevator group, and the control parameters set in each operation rule are determined from the future number of occurrences, and the operation rule is determined. And outputting a control parameter.
Effect of the invention
[0014]
　According to one aspect of the present invention, by realizing optimum elevator control, dissatisfaction of persons involved in the elevator can be reduced. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.
Brief description of the drawings
[0015]
FIG. 1A is a block diagram showing an overall configuration of a group management elevator control system according to an embodiment of the present invention.
FIG. 1B is a block diagram showing a hardware configuration of the analysis server according to the embodiment of the present invention.
FIG. 2 is an explanatory diagram showing the processing of the group management elevator control system of the embodiment of the present invention and the relationship of data.
[Fig. 3] Fig. 3 is a sequence diagram showing an outline of predicted number of people and prediction of destination floor in processing of the group management elevator control system according to the embodiment of the present invention.
FIG. 4 is a sequence diagram showing an outline of effective rule/parameter selection of processing of the group management elevator control system according to the embodiment of the present invention.
FIG. 5 is a flowchart showing a process of an occurrence number estimation model generation unit according to the embodiment of the present invention.
FIG. 6 is a flowchart showing a process of an occurrence number estimation unit according to the embodiment of the present invention.
FIG. 7 is a flowchart showing a process of an occurrence number prediction unit according to the embodiment of the present invention.
FIG. 8 is a flowchart showing a process of a destination floor estimating unit according to the embodiment of the present invention.
FIG. 9 is a flowchart showing a process of a destination floor prediction unit according to the embodiment of the present invention.
FIG. 10 is a flowchart showing a process of a control selector unit according to the embodiment of the present invention.
FIG. 11 is a flowchart showing processing of a rule/parameter evaluation unit according to the embodiment of the present invention.
FIG. 12 is an explanatory diagram of building basic information held by the analysis server according to the embodiment of this invention.
FIG. 13 is an explanatory diagram of a random seed held by the analysis server according to the embodiment of this invention.
FIG. 14 is an explanatory diagram of the number of people getting on and off held by the analysis server according to the embodiment of the present invention.
FIG. 15 is an explanatory diagram of an elevator operation log held by the analysis server of the embodiment of the present invention.
FIG. 16 is an explanatory diagram of external information (weather) held by the analysis server according to the embodiment of the present invention.
FIG. 17 is an explanatory diagram of external information (camera) held by the analysis server according to the embodiment of the present invention.
FIG. 18 is an explanatory diagram of external information (building information) held by the analysis server according to the embodiment of the present invention.
FIG. 19 is an explanatory diagram of occurrence number estimation input held by the analysis server according to the embodiment of the present invention.
FIG. 20 is an explanatory diagram of a model for estimating the number of generated persons held by the analysis server according to the embodiment of the present invention.
FIG. 21 is an explanatory diagram of a result of estimation of the number of generated persons held by the analysis server according to the embodiment of the present invention.
[Fig. 22] Fig. 22 is an explanatory diagram of a predicted number of occurrences held by the analysis server of the embodiment of the present invention.
[Fig. 23] Fig. 23 is an explanatory diagram of a predicted number 2 of generated persons held by the analysis server according to the embodiment of the present invention.
FIG. 24 is an explanatory diagram of time zone-specific destination floor estimation held by the analysis server according to the embodiment of the present invention.
FIG. 25 is an explanatory diagram of destination floor prediction results by time zone held by the analysis server according to the embodiment of the present invention.
FIG. 26 is an explanatory diagram of a rule/control template held by the analysis server according to the embodiment of this invention.
FIG. 27 is an explanatory diagram of a KPI list held by the analysis server according to the embodiment of this invention.
FIG. 28 is an explanatory diagram of simulation inputs and results held by the analysis server according to the embodiment of the present invention.
FIG. 29 is an explanatory diagram of valid rules/parameters held by the analysis server according to the embodiment of this invention.
FIG. 30 is an explanatory diagram of a subdivided list of valid rules/parameters held by the analysis server according to the embodiment of this invention.
FIG. 31 is an explanatory diagram of a rule/parameter list held by the analysis server according to the embodiment of this invention.
FIG. 32 is an explanatory diagram of a building individualization report output by the analysis server according to the embodiment of this invention.
MODE FOR CARRYING OUT THE INVENTION
[0016]
　Next, embodiments of the present invention will be described in detail with reference to the drawings, but the present invention is not limited to the following embodiments, and various modifications and applications within the technical concept of the present invention. Is also included in the range. An embodiment according to the present invention will be described below with reference to FIG.
[0017]
　FIG. 1A is a block diagram showing an overall configuration of a group management elevator control system according to an embodiment of the present invention.
[0018]
　The analysis server SA, the client terminal CL, the external information neighboring building information EXN, the external information database EXD, the external information camera EXC, the control panel CA, the basket 1CA1, the basket 2CA2, and the basket 8CA8 are connected to the network NW of open or close. There is.
[0019]
　The analysis server SA configures an elevator analysis system that analyzes the control of the group management elevator. The analysis server SA includes a database SA0, a display unit SA1, a request unit SA2, and an execution unit SA3.
[0020]
　The database SA0 handles input/output data used in the analysis server SA. Specifically, the database SA0 includes information preset in the analysis server SA, information acquired via the network NW, information generated by the processing of the execution unit SA3, and the like. Although omitted in FIG. 1A, the database SA0 includes, for example, building basic information SA00, random seed SA01, boarding/alighting number SA02, elevator operation log SA03, external information (weather) SA04, external information (camera) SA05, external information ( Building information) SA06, occurrence number estimation input SA07, occurrence number estimation model SA08, occurrence number estimation result SA09, occurrence number prediction result SA10, occurrence number prediction result SA11, destination floor estimation SA12 by time zone, destination floor forecast result by time zone SA13, rule/control template SA14, KPI list SA15, simulation input and result SA16, valid rule/parameter SA17, and rule/parameter list SA19 (see FIGS. 2 and 12 to 31).
[0021]
　The execution unit SA3 is a unit that actually executes the analysis, and includes a measurement processing unit SA31, an occurrence number estimation model generation unit SA32, an occurrence number estimation unit SA33, an occurrence number prediction unit SA34, a destination floor estimation unit SA35, and a destination floor estimation unit SA36. , A control selector section SA37 and a rule/parameter evaluation section SA38.
[0022]
　The client terminal CL is a terminal for an administrator to browse the analysis status. External information The neighboring building information EXN, the external information database EXD, the external information camera EXC, and the control panel CA provide information that is not information about the operation of the elevator. These are described as external information. The external information may include publicly available information such as rail traffic data and road condition data in addition to the above. The baskets 1CA1, 2CA2 and 8CA8 are elevator baskets, and the control panel CA is a control device for controlling the baskets 1CA1 to 8CA8.
[0023]
　Although part of the illustration is omitted, in the example of FIG. 1A, eight baskets 1CA1 to 8CA8 are controlled by the control panel CA. The baskets 1CA1 to 8CA8 are, for example, baskets of eight elevator groups which are set in the same building and face the same elevator hall (elevator hall) and which are the targets of group management. However, 8 units are one example, and the present invention can be applied to an elevator group consisting of 2 or more units.
[0024]
　Further, in the present embodiment, the number of people who are generated is the number of people who have appeared in the elevator hall to get on the elevator.
[0025]
　FIG. 1B is a block diagram showing a hardware configuration of the analysis server SA according to the embodiment of this invention.
[0026]
　The analysis server SA is, for example, a computer having an interface (I/F) 101, an input device 102, an output device 103, a processor 104, a main storage device 105, and an auxiliary storage device 106 that are connected to each other.
[0027]
　The interface 101 is connected to the network NW and communicates with the client terminal CL, the external information database EXD, the external information camera EXC, and the control panel CA via the network NW, and acquires the external information neighborhood building information EXN. The input device 102 is a device used by a user of the analysis server SA to input information to the analysis server SA, and may include, for example, at least one of a keyboard, a mouse, a touch sensor, and the like. The output device 103 is a device that outputs information to the user of the analysis server SA, and may include, for example, a display device that displays characters and images.
[0028]
　The processor 104 executes various processes according to a program stored in the main storage device 105. The main storage device 105 is, for example, a semiconductor storage device such as a DRAM, and stores a program executed by the processor 104 and data necessary for the processing of the processor. The auxiliary storage device 106 is a relatively large-capacity storage device such as a hard disk drive or a flash memory, and stores data and the like referred to in the process executed by the processor.
[0029]
　In the main storage device 105 of this embodiment, a measurement processing unit SA31, an occurrence number estimation model generation unit SA32, an occurrence number estimation unit SA33, an occurrence number estimation unit SA34, a destination floor estimation unit SA35, a destination floor included in the execution unit SA3. Programs for implementing the prediction unit SA36, the control selector unit SA37, and the rule/parameter evaluation unit SA38 are stored. Therefore, in the following description, the processing executed by each unit included in the execution unit SA3 is actually executed by the processor 104 according to the program corresponding to each unit stored in the main storage device 105.
[0030]
　The processing of the request unit SA2 may be realized by the processor 104 controlling the interface 101 or the input device 102 according to a program corresponding to the request unit SA2 stored in the main storage device 105. The processing of the display unit SA1 may be realized by the processor 104 controlling the output device 103 according to a program corresponding to the display unit SA1 stored in the main storage device 105.
[0031]
　The auxiliary storage device 106 of this embodiment stores a database SA0. Further, a program corresponding to each unit included in the execution unit SA3 may be stored in the auxiliary storage device 106 and copied to the main storage device 105 as necessary. Further, at least a part of the database SA0 may be copied to the main storage device 105 as needed.
[0032]
　FIG. 2 is an explanatory diagram showing the processing of the group management elevator control system according to the embodiment of the present invention and the relationship of data.
[0033]
　By referring to this figure, the input data and the output data for each process will be clarified. In addition, it is possible to obtain an overview of processing and data. The block in the bold line frame indicates processing, and the block in the thin line frame indicates data. Further, it is desirable that the range surrounded by the solid line is real-time processing, and the range surrounded by the broken line is executed offline.
[0034]
　Specifically, the number-of-occurred-people estimation model generation unit SA32 executes the number-of-occurred-people model processing SP01 (FIG. 5) based on the building basic information SA00 (FIG. 12) and the random seed SA01 (FIG. 13) to generate the number of generated people. The estimation model SA08 (FIG. 20) is output. In the example of FIG. 2, this generated-person model processing SP01 is executed as offline processing SZ1.
[0035]
　The number-of-occurrence-people estimating unit SA33 detects the number of people boarding/alighting SA02 (FIG. 14), elevator operation log SA03 (FIG. 15), external information (weather) SA04 (FIG. 16), external information (camera) SA05 (FIG. 17), external information (building). (Information) SA06 (FIG. 18), building basic information SA00, and the number-of-occurrence-people estimation model SA08, the number-of-occurrence-people estimation process SP02 (FIG. 6) is executed, and the result is input to the number-of-occurrence-people prediction unit SA34. The occurrence number estimation unit SA33 may use external information other than the above, if available.
[0036]
　The number-of-occurrence-people prediction unit SA34 executes the number-of-occurrence-people prediction process SP03 (FIG. 7) based on the result of the number-of-occurrence-people estimation process SP02, and inputs the result to the destination floor prediction unit SA36 and the control selector unit SA37. Further, the result is saved by the saving process SP07 (FIG. 9).
[0037]
　The destination floor estimation unit SA35 executes the destination floor estimation process SP04 (FIG. 8) based on the number of passengers SA02 getting on and off, and inputs the result to the destination floor prediction unit SA36.
[0038]
　The destination floor prediction unit SA36 executes the destination floor prediction process SP05 (FIG. 9) based on the results of the number-of-occurring-people prediction process SP03 and the destination floor estimation process SP04. The result is saved by the saving process SP07.
[0039]
　The control selector unit SA37 controls the control selector SP06( based on the result of the destination floor prediction process SP05, the result of the occurrence number prediction process SP03, and the rule/parameter list SA19 generated by the display/control data generation process SP15 described later. 10) is executed and the result is output to the control panel CA.
[0040]
　In the example of FIG. 2, the number-of-occurred-people estimation process SP02 to the saving process SP07 are executed as a real-time process SZ0.
[0041]
　The rule/parameter evaluation unit SA38, based on the data saved by the saving process SP07, the rule/control template SA14 (FIG. 26), and the KPI list SA15 (FIG. 27), the KPI simulation process SP11, the valid rule/parameter selection SP12. The end determination processing SP13, the effective rule/parameter subdivision processing SP14, and the display/control data generation processing SP15 are executed (FIG. 11). In the process, simulation input and result SA16 (FIG. 28) and effective rule/parameter SA17 (FIG. 29) are generated, and finally rule/parameter list SA19 (FIG. 31) and building individualization report SA20 (FIG. 32) are generated. Is output.
[0042]
　In the example of FIG. 2, the above KPI simulation processing SP11 to display/control data generation processing SP15 are executed as offline processing SZ2.
[0043]
　FIG. 3 is a sequence diagram showing an outline of the predicted number of people and the prediction of the destination floor in the processing of the group management elevator control system according to the embodiment of the present invention.
[0044]
　The sequence diagram of FIG. 3 corresponds to each of data-related (passenger number SA02, elevator operation log SA03, external information (weather) SA04, external information (camera) SA05, etc.), analysis server SA, control panel CA, and client terminal CL. It is expressed using four axes.
[0045]
　The data collection S01 is a process in which an external system such as the external information database EXD periodically transmits data to the analysis server SA. The measurement processing unit SA31 of the execution unit SA3 of the analysis server SA receives these data, performs database registration S02, and stores the received data in the respective tables of the database SA0. The processes of the number-of-occurrence-people estimation model generation unit SA32, the number-of-occurrence-people estimation unit SA33, the number-of-occurrence-people estimation unit SA34, the destination floor estimation unit SA35, the destination floor prediction unit SA36, and the control selector unit SA37 of the execution unit SA3 are periodically executed, The resulting data is stored in each table of the database SA0 by the database registration S03. Finally, the analysis server SA transmits the control parameter selected by the control selector section SA37 to the control panel CA as an input command CA0.
[0046]
　FIG. 4 is a sequence diagram showing an outline of effective rule/parameter selection of processing of the group management elevator control system according to the embodiment of the present invention.
[0047]
　The sequence diagram of FIG. 4 is expressed using the same four axes as in FIG.
[0048]
　In the client terminal CL, a manager inputs S04 in which the manager inputs, for example, KPI and period as manager information. The client terminal CL sends a request command including the input information to the analysis server SA.
[0049]
　The analysis server SA executes the data acquisition S05 to acquire the data transmitted from the client terminal CL. Then, the analysis server SA executes the rule/parameter evaluation unit SA38 to select useful rules and control parameters while acquiring the corresponding data from the database SA0, and uses the results to generate content. The analysis server SA transmits the display data including the generated content to the client terminal CL.
[0050]
　The client terminal CL executes the display process S06 to display the content. An example of the displayed content will be described later with reference to FIG. Since the KPI and period are used during analysis, it is desirable to register in advance.
[0051]
　FIG. 5 is a flowchart showing the processing of the number-of-occurred-people estimation model generation unit SA32 according to the embodiment of the present invention.
[0052]
　The number-of-occurrences model processing SP01 includes number-of-occurrence-people data generation SP010 and number-of-occurrence-people estimation model generation SP011. In the generated-person data generation SP010, the generated-person estimation model generation unit SA32 uses the building basic information SA00 and the random seed SA01 to perform a simulation (second simulation) on the basis of the number of persons getting on and off (that is, the number of persons getting on and off) and the basket. Ask for status.
[0053]
　For example, the number-of-occurrence-people estimation model generation unit SA32 randomly generates a plurality of people who are about to board the elevator in the elevator halls on each floor during the simulation. Specifically, the number-of-occurrence-people estimation model generation unit SA32 uses the random seed SA01 to randomly determine the floor of the elevator hall in which each person appears and the time of appearance. Further, the occurrence number estimation model generation unit SA32 randomly determines the destination floor of each person from the floors that can be selected based on the building basic information SA00.
[0054]
　Then, the number-of-occurrence-people estimation model generation unit SA32 performs a simulation of operating each car according to the determined time at which each person appears, the floor at which the person appears, and the destination floor, and the number of people getting in and out of each car and the car state at each time. And generate them as the occurrence number estimation input SA07. The car status is, for example, the floor where each car is located, the traveling direction of each car (upward or downward), and the number of passengers in each car, and more specifically, the elevator operation log SA03 described later. It may be the same as the registered value. However, although the actually measured value is registered in the elevator operation log SA03, the occurrence number estimation model generation unit SA32 generates the value by simulation.
[0055]
　At this time, the number-of-occurrence-people estimation model generation unit SA32 may execute the simulation in accordance with, for example, any operation rule/control parameter registered in a rule/control template SA14 (FIG. 26) described later.
[0056]
　In the generated number estimation model generation SP011, the generated number estimation model generation unit SA32 performs a two-step process including step 1 and step 2 to generate the generated number estimation model SA08. In step 1, the number-of-occurrence-people estimation model generation unit SA32 specifies the number of people occurring at each time, the number of people getting on and off, and the basket state obtained corresponding to each, from the result of the simulation. Then, in step 2, the occurrence number estimation model generation unit SA32 estimates the number of occurrences from the state of each car, the number of passengers in each car on each floor, and the number of people getting off, that is, the number of occurrences=f (number of people getting in and out, basket The function f that satisfies the condition) is obtained. For example, as will be described later with reference to FIG. 20, multiple regression analysis may be performed by using the number of people getting on and off and the basket state as an explanatory index, and the number of people as a target variable.
[0057]
　At this time, if at least one of the elevator operation log SA03, external information (weather) SA04, external information (camera) SA05, and external information (building information) SA06 (or other external information) is available, those values ​​are set. As an external variable, a function f may be obtained such that the number of people in the event=f (number of people getting on and off, car condition, external variable). When obtaining the function f, it is sufficient to obtain the inverse conversion of the conversion from the number of passengers specified in step 1 to the number of people getting on and off. Also, another method may be used as long as the function f can be obtained.
[0058]
　FIG. 6 is a flowchart showing the processing of the number-of-occurred-people estimation unit SA33 according to the embodiment of the present invention.
[0059]
　In the number-of-occupations estimation process SP02, the number-of-occurrence estimating section SA33 uses the number-of-occurrence estimation model SA08 obtained in FIG. By substituting, the current occurrence number estimation result SA09 is obtained and held in the main storage device 105 or the auxiliary storage device 106. As a result, it is possible to estimate the occurrence status of a person who is about to get on the elevator from the situation of getting on and off the person in each car and the state of the position and traveling direction of each car.
[0060]
　The number of people getting on and off each car can be estimated, for example, from the change in the weight of each car measured by the control panel CA. Further, the position, traveling direction, etc. of each basket depend on the control by the control panel CA. Therefore, according to the number-of-occurrences model processing SP01 and the number-of-occurrence-number estimation process SP02, even if no external information is obtained, the passenger is about to board the elevator based on the information acquired from the elevator itself. It is possible to estimate the situation of occurrence of a person.
[0061]
　Note that the current boarding/alighting number SA02 and the elevator operation log SA03 show that the operation rules/control parameters (that is, the control panel CA controls the respective baskets based on the operation rules/control parameters applied to the elevator when the data included therein are acquired. It may include information for identifying the operation rule/control parameter that has been performed, for example, see FIG. 26. In that case, the number-of-occurrence-people estimation unit SA33 acquires the number-of-occurrence-people estimation model SA08 generated by the number-of-occurrence-people estimation model generation unit SA32 based on the simulation according to the operation rule/control parameter from the current number of passengers boarding/alighting SA02 and the elevator operation log SA03. By substituting the actual number of passengers, the number of people getting off, and the basket state, the current number of generated persons estimation result SA09 is obtained. As a result, highly accurate estimation can be performed.
[0062]
　Further, at this time, if at least one of the elevator operation log SA03, external information (weather) SA04, external information (camera) SA05, and external information (building information) SA06 (or other external information) is available, those are used. You may substitute.
[0063]
　FIG. 7 is a flowchart showing the processing of the number-of-occurred-people prediction unit SA34 according to the embodiment of the present invention.
[0064]
　The number-of-occurring-people prediction process SP03 includes a number-of-occurring-people prediction SP030 and a format conversion SP031.
[0065]
　In the number-of-occurrence-people prediction SP030, the number-of-occurrence-people prediction unit SA34 uses the number-of-occurrence-people estimation result SA09 at each time point (for example, each time period having a predetermined time width) obtained in the process of FIG. The number of people occurring at a time in the future from the time stored in SA09 is predicted, and the result is output as the number-of-occurrence-people prediction result SA10. At this time, if at least one of the elevator operation log SA03, external information (weather) SA04, external information (camera) SA05, and external information (building information) SA06 (or other external information) is available, use them. May be. For example, when the external information (camera) SA05 is available, the number of people (SA057) included in the external information (camera) SA05 may be used instead of the number of occurrences acquired from the occurrence number estimation result SA09. , If the number of occurrences specified from other external information is available, they may be used.
[0066]
　In the format conversion SP031, the number-of-occurrence-people prediction result SA10 obtained in the number-of-occurrence-people prediction SP030 is converted into the occurrence probability of each person per unit time using the Poisson distribution, and the result is output as the number-of-occurrence-people prediction result 2_SA11. .. The KPI simulation described later can be executed using this occurrence probability.
[0067]
　FIG. 8 is a flowchart showing the processing of the destination floor estimation unit SA35 according to the embodiment of this invention.
[0068]
　The destination floor estimation unit SA35 estimates the destination floor by obtaining the tendency of people getting out of the basket for each time zone based on the number of passengers SA02 getting on and off, and outputs the result as the time zone-specific destination floor estimation SA12.
[0069]
　FIG. 9 is a flowchart showing the processing of the destination floor prediction unit SA36 according to the embodiment of the present invention.
[0070]
　The destination floor prediction unit SA36 executes a destination floor prediction process SP05 and a storage process SP07.
[0071]
　In the destination floor prediction processing SP05, the destination floor prediction unit SA36 predicts the destination floor of the generated people, that is, which floor the generated people are going to go. Specifically, the destination floor estimation unit SA36 estimates the number of generated persons 2_SA11 that is the processing result of the number of generated persons prediction unit SA34 of FIG. 7 and the destination floor estimation by time zone that is the processing result of the destination floor estimation unit SA35 of FIG. The destination floor is predicted by multiplying with SA12, and the result is output as the destination floor prediction result by time zone SA13.
[0072]
　The saving process SP07 is a process of saving the number-of-occurred-people prediction result SA10 and the time-slot-specific destination floor prediction result SA13 that have been obtained so far, for example, in the database SA0. The reason for doing this is that a large amount of past data is required when performing offline processing.
[0073]
　FIG. 10 is a flowchart showing the processing of the control selector section SA37 according to the embodiment of the present invention.
[0074]
　The control selector SP06 executed by the control selector unit SA37 is a process of selecting a rule/parameter list that is satisfied by the number-of-occurrence-people prediction result SA10 and the time zone-specific destination floor prediction result SA13. The selected parameter is sent to the control panel CA as the input command CA0.
[0075]
　FIG. 11 is a flowchart showing processing of the rule/parameter evaluation unit SA38 according to the embodiment of this invention.
[0076]
　The rule/parameter evaluation unit SA38 includes a KPI simulation process SP11, a valid rule/parameter selection SP12, an end determination process SP13, a valid rule/parameter subdivision process SP14, and a display/control data generation process SP15.
[0077]
　In the KPI simulation processing SP11, the rule/parameter evaluation unit SA38 changes the operation rule/control parameter using the rule/control template SA14, the KPI list SA15, the time zone-specific destination floor prediction result SA13, and the number-of-occurrences prediction result 2_SA11. While performing the simulation a plurality of times (first simulation), the KPI value is output.
[0078]
　Specifically, the rule/parameter evaluation unit SA38 calculates the destination floor probability of each floor according to the destination floor probability entered in the time zone-specific destination floor estimation SA12 and the occurrence probability of each person entered in the occurrence number prediction result 2_SA11. A simulation is executed in which a person is generated in the elevator hall, and each car is controlled according to the operation rule/control parameter selected from the rule/control template SA14 accordingly. As will be described later, this simulation is executed multiple times while changing the operation rules/control parameters to be applied.
[0079]
　In the valid rule/parameter selection SP12, the rule/parameter evaluation unit SA38 selects a valid rule/parameter from the value substituted into the KPI simulation processing SP11 and the result.
[0080]
　In the end determination processing SP13, the rule/parameter evaluation unit SA38 determines whether or not the improvement effect is seen as a result of the valid rule/parameter selection SP12, and Yes is obtained if the improvement effect is seen, and No if not. Proceed to.
[0081]
　In the effective rule/parameter subdivision processing SP14, the rule/parameter evaluation unit SA38 determines a range for subdividing the more effective feature amount from the result of the effective rule/parameter selection SP12.
[0082]
　The rule/parameter evaluation unit SA38 substitutes the result into the KPI simulation process SP11 to repeat the loop until the effect of improving the result by the end determination process SP13 is seen.
[0083]
　In the display/control data generation process SP15, the rule/parameter evaluation unit SA38 generates a rule/parameter list SA19 and a building individualization report SA20 based on the valid rule/parameter SA17.
[0084]
　FIG. 12 is an explanatory diagram of the building basic information SA00 held by the analysis server SA according to the embodiment of this invention.
[0085]
　The building basic information SA00 is a table in which basic building information is described. The elevator is composed of multiple baskets for each building, and it is called an elevator bank. Since the control is performed for each elevator bank, the table for managing it is the building basic information table (FIG. 12). For example, the car 1CA1 to the car 8CA8 in FIG. 1A belong to one elevator bank. One elevator bank corresponds to an elevator group that is the target of group management by the control panel CA. When there are a plurality of elevator banks in one building, there are a plurality of combinations of control panels and a plurality of baskets.
[0086]
　The building ID (SA000) is identification information (ID) of the building in which the elevator is installed. It is identified by a different ID for each building. The elevator bank ID (SA001) is an ID for distinguishing elevator banks in a building. The bank name (SA002) is the name of the elevator bank. The number of baskets (SA003) is the number of baskets forming the elevator bank. The target floor (SA004) indicates the floor where the baskets forming the elevator bank stop. The latitude (SA005) and the longitude (SA006) are the latitude and the longitude indicating the position where the elevator bank is, respectively. If the area of ​​the elevator bank is large, it may be the latitude and longitude of its center of gravity. Further, it is sufficient that there is information that indicates the position of the elevator bank in absolute coordinates with the entire earth, and values ​​other than latitude and longitude may be used. The building name (SA007) is the official name of the building where the elevator is located.
[0087]
　FIG. 12 shows an example, and if there is data required as basic information of a building during analysis, the basic building information SA00 can be changed to add the data.
[0088]
　FIG. 13 is an explanatory diagram of the random seed SA01 held by the analysis server SA according to the embodiment of this invention.
[0089]
　The random seed SA01 is a table in which a seed used when generating a random number is described. The random seed No (SA010) is the ID of the random seed. It is identified by a different ID for each random seed.
[0090]
　Random seed (SA011) is a random seed value. By using this table, when a random seed No is specified, the corresponding value can be referenced.
[0091]
　FIG. 13 shows an example, and when generating random numbers, if there is necessary data, the random seed SA01 can be changed so as to add the data.
[0092]
　FIG. 14 is an explanatory diagram of the number of boarding/alighting passengers SA02 held by the analysis server SA according to the embodiment of this invention.
[0093]
　The boarding/alighting number SA02 is a table showing the number of passengers and the number of passengers getting on and off each basket for each floor by the actual elevator.
[0094]
　The building ID (SA020) is an ID for identifying a building. The elevator bank ID (SA021) is an ID for identifying each of a plurality of elevator banks in the building. The date (SA022) is the date showing the operating status of the elevator. Time (SA023) is a time indicating the operation status of the elevator.
[0095]
　The day of the week (SA024) is the day of the week on which the operation status of the elevator is shown. The time width (SA025) is a time width in which the operation status of the elevator is totaled. Basket 1 (SA026) indicates that one basket belonging to the elevator bank identified by the elevator bank ID (SA021) is identified. The floor (SA027) is a floor in which the basket 1 (SA026) exists in the time zone specified by the date (SA022), the time (SA023), the day of the week (SA024), and the time width (SA025). The number of persons in the basket (SA028) is the number of persons who were in the basket during the time period specified by the date (SA022), the time (SA023), the day of the week (SA024) and the time width (SA025) (that is, the number of persons in the basket). ).
[0096]
　As for the number of passengers (SA02A) in the upward direction (SA029), the car 1 (SA026) is in the upward direction in the time zone specified by the date (SA022), the time (SA023), the day of the week (SA024), and the time width (SA025). It shows the number of people (that is, the number of passengers) who got into the car when facing the car. As for the number of descending people (SA02B) in the upward direction (SA029), the basket 1 (SA026) is in the upward direction in the time zone specified by the date (SA022), the time (SA023), the day of the week (SA024), and the time width (SA025). It shows the number of people (that is, the number of people getting off) from the car when facing the car.
[0097]
　As for the number of passengers (SA02D) in the downward direction (SA02C), the basket faces downward in the time zone specified by the date (SA022), the time (SA023), the day of the week (SA024), and the time width (SA025). The number of passengers at that time is shown. The number of people (SA02E) descending in the downward direction (SA02C) is downward in the time zone specified by the date (SA022), time (SA023), day of the week (SA024), and time width (SA025). It shows the number of people getting off at that time.
[0098]
　The boarding/alighting number SA02 includes information on all the baskets forming the elevator bank. The information about the basket 1 (SA026) shown in FIG. 14 is one of them. Although not shown in FIG. 14, data regarding other cars are also stored as the number of passengers SA02.
[0099]
　The timing at which data is entered in the boarding/alighting passengers SA02 is for each event (for example, when there is an actual change) or for each predetermined cycle (for example, every 1 millisecond, every one second, every one minute, etc.). But it's okay. The date and time when the information is actually entered may be indicated by the date (SA022), the time (SA023) and the day of the week (SA024). When the information is entered every predetermined period, the period may be entered as the time width (SA025). Further, it is not necessary that all the data specified in this table be stored.
[0100]
　For example, the first row of the table in FIG. 14 is the elevator bank ID “01” of the building identified by the building ID “B001” during 5 minutes starting from 10:00:01 am on Tuesday, June 27, 2017. The car 1 (SA026) belonging to the elevator bank identified by No. 1 or more stops on the 3rd floor at least once, and the number of people in the car (SA028) at that time is 10 people. SA02A) and the number of people getting off (SA02bB are 15 people and 1 person respectively), and the number of passengers (SA02D) and the number of people getting off (SA02E) when stopping while moving downward were 0 person and 10 people, respectively. The number of people in the car (SA028) is the number of people after getting on and off the stopped floor.These numbers are those of the car 1 (SA026) stopped on the 3rd floor multiple times in the above 5 minutes. In this case, the total number of those times may be entered, or the number of people stopped once on the 3rd floor may be entered, and the car 1 (SA026) concerned is different in the same 5 minutes. If the floor is stopped more than once, the same information as above is entered in the table for that floor as well.
[0101]
　FIG. 14 shows an example, and when expressing the number of people getting on and off by floor, if there is necessary data, the number of passengers SA02 can be changed so as to add the data.
[0102]
　FIG. 15 is an explanatory diagram of the elevator operation log SA03 held by the analysis server SA according to the embodiment of this invention.
[0103]
　The elevator operation log SA03 is a table showing an operation log of an actual elevator. This table can store both data aggregated for each elevator bank and data for cars belonging to the elevator bank.
[0104]
　The building ID (SA030) is an ID that identifies a building. The elevator bank ID (SA031) is an ID that identifies a plurality of elevator banks in the building. The date (SA032) is the date showing the operating status of the elevator. Time (SA033) is a time indicating the operation status of the elevator.
[0105]
　The day of the week (SA034) is the day of the week on which the operation status of the elevator is shown. The time width (SA035) is a time width in which the operation status of the elevator is totaled. The long wait rate (SA036) is the waiting time (that is, the person who called the basket) that occurred in the elevator bank in the time zone specified by the date (SA032), the time (SA033), the day of the week (SA034), and the time width (SA035). Shows the ratio of the waiting time of a predetermined length (for example, 60 seconds) or more of the time waiting until the basket arrives. The predetermined length can be changed by designating in advance.
[0106]
　The number of basket calls (SA037) is the number of times the basket call button is pressed in the elevator bank during the time zone specified by the date (SA032), time (SA033), day of the week (SA034) and time width (SA035). .. Traffic flow mode (SA038) is an elevator bank operation mode.
[0107]
　The long wait rate (SA036), the number of basket calls (SA037), and the traffic flow mode (SA038) are values ​​aggregated for each elevator bank. However, if there is necessary data, change the above information and add information other than the above. can do.
[0108]
　Basket 1 (SA039) indicates that one basket belonging to the elevator bank ID (SA031) is identified. The floor (SA0A) is the position (floor) where the basket 1 (SA039) was present at the time specified by the date (SA032), the time (SA033), and the day of the week (SA034). The direction (SA03B) is the direction in which the car 1 (SA039) is proceeding to the time point specified by the date (SA032), the time (SA033), and the day of the week (SA034). For example, the upper side indicates that the vehicle is moving upward, and the lower side indicates that the vehicle is moving downward.
[0109]
　The state (SA03C) indicates the state of the basket 1 (SA039) at the time point specified by the date (SA032), the time (SA033), and the day of the week (SA034). For example, "motion" indicates that the basket 1 (SA039) was actually moving, and "stop" indicates that it was stopped. The number of passengers (SA03D) indicates the number of passengers in the car 1 (SA039) at the time specified by the date (SA032), time (SA033), and day of the week (SA034).
[0110]
　The elevator operation log SA03 includes information about all the baskets that make up the elevator bank. The basket 1 (SA039) shown in FIG. 15 is one of them. Although omitted in FIG. 15, data regarding other cars are also stored as the elevator operation log SA03.
[0111]
　The timing at which data is entered in the elevator operation log SA03 is for each event (for example, when there is an actual change) or for each predetermined cycle (for example, every 1 millisecond, every one second, every one minute, etc.). ) May be. The date and time when the information is actually entered may be indicated by the date (SA032), the time (SA033) and the day of the week (SA034). When the information is entered every predetermined period, the period may be entered as the time width (SA035). Further, it is not necessary that all the data specified in this table be stored.
[0112]
　FIG. 15 is an example, and when expressing the operation log by the elevator, if there is necessary data, the elevator operation log SA03 can be changed to add the data.
[0113]
　FIG. 16 is an explanatory diagram of external information (weather) SA04 held by the analysis server SA according to the embodiment of this invention.
[0114]
　The external information (weather) SA04 is a table in which data on weather, which is one of the external information, is collected.
[0115]
　The external information ID (SA040) is an identification ID of external information. The date (SA041) is the date when the external information is acquired. Time (SA042) is the time when the external information is acquired. The day of the week (SA043) is the day of the week when the external information is acquired. The place (SA044) is the place where the external information is acquired. Latitude (SA045) is the latitude at which the external information is acquired. The longitude (SA046) is the longitude at which the external information is acquired. Weather (SA047), temperature (SA048), and rainfall (SA049) are the weather, temperature, and rainfall at the location specified by the location (SA044) at the time specified by the date (SA041) and time (SA042), respectively. is there.
[0116]
　The data is written in the external information (weather) SA04 at each event (for example, when there is an actual change) or every predetermined period (for example, every 1 millisecond, every one second, every one minute). Etc.) The date and time when the data was actually entered and the place where the data was acquired may be indicated by date and time (SA041), time (SA042) and place (SA044). Further, it is not necessary that all the data specified in this table be stored.
[0117]
　FIG. 16 shows an example, and when expressing data related to weather, which is one of the external information, if there is necessary data, change the external information (weather) SA04 so as to add the data. can do.
[0118]
　FIG. 17 is an explanatory diagram of external information (camera) SA05 held by the analysis server SA according to the embodiment of this invention.
[0119]
　The external information (camera) SA05 is a table in which data related to what is recognized by the measurement by the camera, which is one of the external information, is collected.
[0120]
　The external information ID (SA050) is an identification ID of external information. The date (SA051) is the date when this information was acquired. The time (SA052) is the time when this information was acquired. The day of the week (SA053) is the day of the week when this information was acquired. The building ID (SA054) is an ID for identifying the building from which this information is acquired. The floor (SA055) is the floor from which this information is acquired. The installation location (SA056) is the location where the camera is installed to acquire this information.
[0121]
　The number of people (SA057) is the number of people detected by the camera installed in the place specified by the installation place (SA056) at the time specified by the date (SA051) and the time (SA052). Children (SA058), adults (SA059), men (SA05A), women (SA05B), wheelchairs (SA05C) and trolleys (SA05D) are installed at the times specified by the date (SA051) and time (SA052), respectively. The number of children, adults, men, women, wheelchairs and trolleys detected by cameras installed at the location specified by the location (SA056). In this way, not only the total number of persons but also a breakdown for each person attribute (for example, age group and sex) and an object other than the person can be detected.
[0122]
　Anger (SA05E) is detected by the camera based on the result detected by the camera installed at the location specified by the installation location (SA056) at the time specified by the date (SA051) and the time (SA052). It is the number of people who were determined to be angry among the number of people. In this way, not only the number of persons but also the emotions of the person can be detected from the face and the behavior by the camera, and the number of persons in which the specific emotion is detected can be counted.
[0123]
　The data may be written in the external information (camera) SA05 at each event (for example, when there is an actual change), or at a predetermined cycle (for example, 1 millisecond, 1 second, 1). Every minute). The date and time when the data is actually entered and the installation location of the camera from which the data is acquired may be indicated by date (SA051), time (SA052), and installation location (SA056). Further, it is not necessary that all the data specified in this table be stored.
[0124]
　FIG. 17 is an example, and when expressing the data related to what is recognized by the measurement by the camera, which is one of the external information, if necessary data is added, the external data is added. The information (camera) SA05 can be changed.
[0125]
　FIG. 18 is an explanatory diagram of external information (building information) SA06 held by the analysis server SA according to the embodiment of this invention.
[0126]
　The external information (building information) SA06 is a table in which data regarding a building, which is one of the external information, is collected.
[0127]
　The external information ID (SA060) is an identification ID of external information. The building ID (SA061) is an ID for identifying the building from which this information is acquired. The date (SA062) is the date when this information was acquired. The time (SA063) is the time when this information was acquired. The day of the week (SA064) is the day of the week when this information was acquired. The east side of the third floor (SA065) indicates the floor (third floor) from which this information has been acquired, and the area (east side) from which the main information has been acquired among the areas obtained by dividing the floor. Stores the values ​​aggregated for each floor and area. Floors and areas can be added arbitrarily, and when added, the data aggregated on the floors and areas can be stored in the same manner as on the east side of the third floor (SA065).
[0128]
　The amount of electricity used (SA066) and the amount of water used (SA067) are the amount of electricity and the amount of water used on the east side of the third floor (SA065) at the time points specified by the date (SA062) and the time (SA063), respectively. The temperature (SA068) and the humidity (SA069) are the temperature and the humidity on the east side of the third floor (SA065) at the time points specified by the date (SA062) and the time (SA063), respectively. The number of stayers (SA06A) is the number of stayers on the east side of the third floor (SA065) at the time point specified by the date (SA062) and the time (SA063).
[0129]
　The data may be written in the external information (building information) SA06 at each event (for example, when there is an actual change) or at a predetermined cycle (for example, every 1 millisecond, every 1 second). (Every minute, etc.). The date and time when the data is actually entered and the place where the data is obtained may be indicated by the date (SA062), the time (SA062), and the east side of the third floor (SA065). Further, it is not necessary that all the data specified in this table be stored.
[0130]
　FIG. 18 shows an example, and when expressing data regarding a building, which is one of the external information, if there is necessary data, the external information (building information) SA06 is added so that the data is added. Can be changed.
[0131]
　FIG. 19 is an explanatory diagram of the occurrence number estimation input SA07 held by the analysis server SA according to the embodiment of this invention.
[0132]
　The number-of-occurrence-people estimation input SA07 is a table that stores data generated by the number-of-occurrence-people data generation SP010 in the number-of-occurrence-people model processing SP01. The generated data includes the number of people who have occurred at each floor, the number of people who get on and off by car, and the car status.
[0133]
　The occurrence number estimation input ID (SA070) is an ID for identifying the occurrence number estimation input value. The time (SA071), the day of the week (SA072), and the time width (SA073) are the time, the day of the week, and the time width generated by the generated person data generation SP010, respectively. The number of generated persons (SA074) is the number of generated persons generated by the generated person number data generation SP010. The number of people is required for each floor. In FIG. 19, the number of persons on the third floor is shown on the third floor (SA075). Although not shown in FIG. 19, the numbers of people on other floors are also entered in the same manner. The number of generated persons can be generated for each floor, each area, and each elevator hall, and in that case, the generated number of generated persons is stored in the number of generated persons (SA074).
[0134]
　The number of people getting in and out by car (SA076) is the number of people getting in and out by car generated by the generated-person data generation SP010 in the time zone specified by the time (SA071), the day of the week (SA072), and the time width (SA073). The number of people getting on and off by car (SA076) is calculated for each car. In FIG. 19, the number of people in and out of the car 1 by car is shown as car 1 (SA077). The information about the number of passengers getting on and off of the car 1 is stored in the car 1 (SA077), the floor (SA078) is the floor where the car is present, and the car moving upward in the upward direction (SA079) is stopped on the floor. The number of passengers and the number of passengers when the car is moving down, and the downward direction (SA07A) is the number of passengers and the number of passengers when the car moving downward is stopped on the floor. In addition to the above, information about the car 1 can also be stored in the car 1 (SA077). As for the number of people by boarding/alighting by car (SA076), it is possible to store information about a car other than the car 1 as well.
[0135]
　The car state (SA07B) stores data on the car state, and the data on car 1 is stored in the car 1 (SA07C). The floor is a floor in which the basket 1 (SA07C) exists in the time zone specified by the time (SA071), the day of the week (SA072), and the time width (SA073). The direction is the direction in which the car 1 (SA07C) is advancing in the time zone specified by the time (SA071), the day of the week (SA072), and the time width (SA073). For example, “up” indicates going up, and “down” indicates going down.
[0136]
　The state shows the state of the basket 1 (SA07C) in the time zone specified by the time (SA071), the day of the week (SA072), and the time width (SA073). For example, “motion” indicates that the object is actually moving, and “stop” indicates that the object is stopped. The number of passengers is the number of passengers in the car 1 (SA07C) during the time period specified by the time (SA071), the day of the week (SA072), and the time width (SA073). In addition to the above, information about the state of the car 1 can be stored in the car 1 (SA07C). The car status (SA07B) can store information about the status of the car other than the car 1 as well.
[0137]
　Data may be written in the occurrence number estimation input SA07 at each event (for example, when there is an actual change), or at a predetermined cycle (for example, every 1 millisecond, every 1 second, 1 minute). It may be). The date and time of actual entry may be indicated by the time (SA071) and the day of the week (SA072). Further, it is not necessary that all the data specified in this table be stored.
[0138]
　FIG. 19 shows an example, and when representing the data generated by the generated-person data generation SP010, if there is necessary data, the generated-person estimation input SA07 is changed to add the data. can do.
[0139]
　FIG. 20 is an explanatory diagram of the number-of-occurrence-people estimation model SA08 held by the analysis server SA according to the embodiment of this invention.
[0140]
　The number-of-occurrence-people estimation model SA08 is a table that stores data generated in the number-of-occurrence-people estimation model generation SP011 in the number-of-occurrence-people model processing SP01. The generated data is the function f when “occurrence number=f (number of people getting on/off, car status, external information)”. The number of people in attendance, the number of people in boarding/alighting, and the basket status are acquired as the number of occurrences estimation input SA07, and external information is acquired from external information (weather) SA04, external information (camera) SA05, and external information (building information) SA06. ..
[0141]
　The occurrence number estimation ID (SA080) is an ID for identifying the occurrence number estimation model. The floor (SA081) is the floor targeted by the generated estimation model. The direction (SA082) is the target direction of the generated estimation model. The time (SA083) is the time targeted by the generated estimation model. The day of the week (SA084) is the day of the week for which the generated estimation model is targeted. The time width (SA082) is the time width that is the target of the generated estimation model.
[0142]
　The coefficient of the function f is stored in the subsequent columns. Occurrence number estimation input SA07 number of persons by boarding/alighting by car (SA076), basket state (SA07B), or one selected from external information (weather) SA04, external information (camera) SA05, and external information (building information) SA06 The data of one or more items is used as the feature amount. Then, the coefficient of the feature amount can be obtained by performing a multiple regression analysis using the feature amount as an explanation index and the number of occurrences (SA074) as the target index. The boarding/alighting passenger coefficient 1 (SA085), the basket state coefficient 1 (SA087), and the external variable coefficient 1 (SA088) are the coefficient of the feature amount obtained by the analysis. Since the coefficient is obtained for each feature amount, it is desirable to store the coefficient for each feature amount.
[0143]
　An analysis method other than the multiple regression analysis may be used as a method for generating a model for estimating the number of people who have occurred.
[0144]
　The data may be written in the occurrence number estimation model SA08 at each event (for example, when there is an actual change) or at a predetermined cycle (for example, 1 millisecond, 1 second, 1 minute). It may be). The date and time of actual entry may be indicated by the time (SA083) and the day of the week (SA084). Further, it is not necessary that all the data specified in this table be stored.
[0145]
　FIG. 20 is an example, and when representing the model generated in the generation number estimation model generation SP011, if there is necessary data, the generation number estimation model SA08 is added so as to add the data. Can be changed.
[0146]
　FIG. 21 is an explanatory diagram of the number-of-occurrence-people estimation result SA09 held by the analysis server SA according to the embodiment of this invention.
[0147]
　The occurrence number estimation result SA09 is a table that stores data generated in the occurrence number estimation SP020 in the occurrence number estimation processing SP02. In the number-of-occurrence-people estimation SP020, the stored number-of-occurrence-people estimation model (function f), the number of people getting on and off at the current time, the basket state, and external variables are input, and the number of people generated by each floor is estimated. The result is stored in the number-of-occurrence-people estimation result SA09 in FIG.
[0148]
　The occurrence number estimation ID (SA090) is an ID for identifying the occurrence number estimation. The date (SA092) is the date on which the number of occurrences was estimated. The time (SA093) is the time at which the number of occurrences was estimated. The day of the week (SA094) is the day of the week on which the number of occurrences is estimated. The time width (SA092) is a time width in which the number of people who have occurred is estimated. The floor (SA093) is the floor in which the number of people who have occurred is estimated. The place (SA094) is the place where the number of occurrences is estimated. The number of occurrence (SA095) is the estimated number of occurrence.
[0149]
　Data may be written in the occurrence number estimation result SA09 at each event (for example, when there is an actual change), or at a predetermined cycle (for example, 1 millisecond, 1 second, 1 minute). It may be). The date and time when the information is actually entered may be indicated by the date (SA092), the time (SA092), and the day of the week (SA093). Further, it is not necessary that all the data specified in this table be stored.
[0150]
　FIG. 21 shows an example, and when expressing the number of people generated in the number-of-occurrence-people estimation SP020, if there is necessary data, change the number-of-occurrence-people estimation result SA09 so as to add the data. can do.
[0151]
　FIG. 22 is an explanatory diagram of the number-of-occurrence-people prediction result SA10 held by the analysis server SA according to the embodiment of this invention.
[0152]
　The number-of-occurrence-people prediction result SA10 is a table storing data generated in the number-of-occurrence-people estimation SP020 in the number-of-occurrence-people prediction process SP03. In the number-of-occurrence-people prediction SP030, the number-of-occurrence-people prediction unit SA34 performs a process of estimating the number of people in the future using the number-of-occurrence-people estimation result SA09 obtained in the number-of-occurrence-people estimation process SP02 and external information. The result is stored in the generated person prediction result SA10 in FIG.
[0153]
　As described with reference to FIG. 7, the input of the number-of-occurrence-people prediction SP030 is the time width used for analysis (for example, the past 10 minutes), the number-of-occurrence-people estimation result SA09, external information (weather) SA04, external information (camera). ) SA05 and external information (building information) SA06, and the output is the number of future generations.
[0154]
　An AR model (autoregressive model) or the like is used as a method for predicting the number of people who have occurred, but an analysis method other than the AR model may be used.
[0155]
　The occurrence number prediction ID (SA100) is an ID for identifying the performed occurrence number prediction. The date (SA101), the time (SA102), and the day of the week (SA103) are the date, time, and day of the week to be analyzed (that is, at the time of analysis), respectively. The predicted time (SA104) is the time at which the analysis target is predicted (that is, the number of people who have occurred at that time is predicted). The time width (SA105) is the time width of the analysis target. The floor (SA106) is the floor to be analyzed. The place (SA107) is the place to be analyzed. The number of occurrences (SA108) is the number of occurrences at the time when the analysis target is predicted.
[0156]
　For example, the first line of the generated person number prediction result SA10 in FIG. 22 indicates that the process of predicting the number of generated persons on the elevator floor on the third floor within 5 minutes from 10:06:01 on Tuesday, June 27, 2017 is the same day. It is executed at 10:01:01, and as a result, it is predicted that the number of occurrences is 12 people.
[0157]
　The timing of substituting the occurrence number prediction result SA10 may be for each event (for example, when there is an actual change) or for each predetermined cycle (for example, every 1 millisecond, every one second, every one minute). Etc.) The date and time when the information is actually entered may be indicated by the date (SA101), the time (SA102), and the day of the week (SA103). Further, it is not necessary that all the data specified in this table be stored.
[0158]
　FIG. 22 shows an example, and when expressing the occurrence number prediction in the occurrence number prediction SP030, if there is necessary data, change the occurrence number prediction result SA10 so as to add the data. You can
[0159]
　FIG. 23 is an explanatory diagram of the generated person prediction result 2_SA11 held by the analysis server SA according to the embodiment of this invention.
[0160]
　The number-of-occurrence-people prediction result 2_SA11 is a table that stores a format conversion of the number-of-occurrence-people prediction generated in the format conversion SP031 in the number-of-occurrence-people prediction process SP03.
[0161]
　In the format conversion SP031, the number-of-occurrence-people prediction unit SA34 uses the number-of-occurrence-people prediction result SA10 to obtain the occurrence probability for each number of people per unit time using the Poisson distribution. The result is stored in the generated person prediction result 2_SA11 in FIG.
[0162]
　The formula of Poisson distribution is as shown in the following formula (1). The probability P(k) of occurrence of k or more persons can be obtained by substituting the number of persons for each floor into λ of the equation (1).
[0163]
[Number 1]

[0164]
　The above example is a method of obtaining the occurrence probability for each occurrence number based on the assumption that the probability distribution of the occurrence number follows Poisson distribution. However, an analysis method other than the method using the Poisson distribution may be used to obtain the probability of the number of occurrences.
[0165]
　The occurrence number prediction ID (SA110) is an ID for identifying the performed occurrence number prediction. The date (SA111), time (SA112), and day of the week (SA113) are the date, time, and day of the week to be analyzed (that is, at the time of analysis), respectively. The prediction time (SA114) is the time at which the analysis target is predicted (that is, the occurrence probability at that time is predicted). The time width (SA115) is the time width of the analysis target. The floor (SA116) is the floor to be analyzed. The place (SA117) is the place to be analyzed. The probability of occurrence of one or more (SA118) is the probability of occurrence of one or more persons per unit time. The probability of occurrence of 2 or more (SA119) is the probability of occurrence of 2 or more persons per unit time. The time width (SA115) may be used as the unit time.
[0166]
　For example, the top row of the generated person prediction result 2_SA11 of FIG. 23 shows an example corresponding to the prediction result entered in the top row of the generated person prediction result SA10 of FIG. That is, the head row of the generated person prediction result 2_SA11 in FIG. 23 shows that the number of persons "12" predicted to occur on the elevator floor on the third floor causes one or more persons on the elevator floor on the third floor per unit time. The probability is 90%, and the probability that two or more persons will occur is predicted to be 75%. Although not shown in FIG. 23, similarly, the probability of occurrence of 3 or more persons, the probability of occurrence of 4 or more persons, etc. are also calculated and entered in the occurrence number prediction result 2_SA11.
[0167]
　FIG. 23 shows an example, and when expressing the occurrence number prediction in the format conversion SP031, if there is necessary data, it is possible to change the occurrence number prediction result 2_SA11 so as to add the data. it can.
[0168]
　FIG. 24 is an explanatory diagram of the destination floor estimation SA12 by time zone held by the analysis server SA according to the embodiment of this invention.
[0169]
　The destination floor estimation SA12 by time zone is a table that stores data generated by the destination floor estimation processing SP04. In the destination floor estimation processing SP04, the destination floor estimation unit SA35 generates a model for estimating the destination for each time zone using the number of people getting on and off by floor (SA02). Specifically, the destination floor estimation unit SA35 counts the number of people getting off by floor for each time zone, and obtains the tendency of the number of people getting off by floor. Then, it is converted into an estimated value with the whole as 100%. The result is stored in the destination floor estimation SA12 by time zone in FIG.
[0170]
　The destination floor estimation ID (SA120) is an ID for identifying the performed destination floor estimation. The date (SA121), the time (SA122), the day of the week (SA123), and the time width (SA124) are the date, time, day of the week, and time width of the analysis target, respectively. The boarding floor (SA125) is the floor on which the passenger gets on. The direction (SA126) is the direction in which the car advances. The destination floor (SA127) is the floor where the passenger gets off. Estimated values ​​for the floor where the elevator is stopped as 100% are entered.
[0171]
　For example, the first line of Fig. 24 shows that 10% of the people who got on the 3rd floor of the car going up from the 3rd floor got off at the 26th floor in the 60 minutes from 10:01:01 on Tuesday, June 27, 2017. , Another 10% got off at the 27th floor, which is estimated from the number of passengers SA02 getting on and off. In FIG. 24, the proportion of persons who get off on the other floors is omitted, but the sum of the proportions calculated for all floors that can be destinations for persons who have boarded from the third floor is 100%. The ratio is similarly calculated for destination floors from other floors. In the present embodiment, these ratios are used as the destination floor probability, which is the probability that the destination floor of the person who appears at the landing on each floor becomes that floor.
[0172]
　If it is possible to determine whether the person riding in the basket on each floor and the person leaving the basket on each floor are the same person based on external information (camera) SA05 or the like, based on the determination result. The person who got on from each floor, such as, for example, 10% of the persons who got on the 26th floor out of the persons who got on the 3rd floor, are identified based on which floor each person who got on each floor got off You can calculate the percentage of your destination floor. However, when such external information is not available, for example, when the number of people getting on and off at each floor cannot be identified, for example, by estimating the number of people getting on and off each floor from the weight of the basket, an approximation based on some assumption is made. The percentage of destination floors may be calculated.
[0173]
　For example, the number of persons who have descended on each floor during the time period specified by the date (SA121), the time (SA122), the day of the week (SA123), and the time width (SA124) is totaled, and the number of persons who have descended on a floor other than the third floor is calculated. The ratio of the number of people who got off on the 26th floor to the total number of the people who got off on the 26th floor among the people who got on the 3rd floor (that is, the probability that the destination floor of the person who got on the 3rd floor is the 26th floor) ). In that case, the ratio of the person who got off on the other floor and the ratio of the person who got on the other floor and got off on each floor are calculated by the same method.
[0174]
　FIG. 24 shows an example. When expressing the destination floor estimation in the destination floor estimation processing SP04, if there is necessary data, the destination floor estimation SA12 by time zone is added so that the data is added. Can be changed.
[0175]
　FIG. 25 is an explanatory diagram of the destination floor prediction result SA13 by time zone held by the analysis server SA according to the embodiment of this invention.
[0176]
　The time zone-specific destination floor prediction result SA13 is a table that stores data generated in the time zone-specific destination floor prediction SP051 in the destination floor prediction processing SP05. In the destination floor prediction SP051 by time zone, the destination floor prediction unit SA36 uses the destination floor estimation SA12 by time zone and the number-of-occurrence-number prediction result 2_SA11 as input data, and combines them to determine which floor the generated people belong to. Can predict whether to visit. Specifically, the occurrence probability of the predicted occurrence time for each floor may be multiplied by the destination floor estimation at the same time.
[0177]
　The time zone destination floor prediction method described above is an example, and other methods may be used. The result is stored in the destination floor prediction result SA13 by time zone in FIG.
[0178]
　The destination floor prediction ID (SA130) is an ID for identifying the performed destination floor prediction. The date (SA131), time (SA132), and day of the week (SA133) are the date, time, and day of the week to be analyzed (that is, at the time of analysis), respectively. The prediction time (SA134) is the time at which the analysis target is predicted (that is, the occurrence probability at that time is predicted). The time width (SA135) is the time width of the analysis target. The passenger floor (SA136) is the passenger floor to be analyzed. The destination floor (SA137) is the destination floor to be analyzed. The direction (SA138) is the direction in which the basket to be analyzed advances. The probability of occurrence of one or more persons (SA139) is the probability of occurrence of one or more persons per unit time. The probability of occurrence of two or more persons (SA13A) is the probability of occurrence of two or more persons per unit time. The time width (SA135) may be used as the unit time.
[0179]
　For example, the top row of the time zone-specific destination floor prediction result SA13 of FIG. 25 is the prediction result entered in the top row of the number-of-occurrence-person prediction result 2_SA11 of FIG. 23 and the time zone-specific destination floor estimation SA12 of FIG. An example corresponding to the estimation result entered in the first row is shown. That is, the top row of the destination floor prediction result SA13 by time zone of FIG. The probability of occurrence is 9%, and the probability of occurrence of two or more persons is predicted to be 7.5%.
[0180]
　In this example, “9%” corresponds to “90%”, which is the probability (SA118) of one or more occurrences in the first row of FIG. 23, and 26th floor of the destination floor (SA127) of the first row in FIG. It is obtained by multiplying the value "10%". “7.5%” is a value corresponding to 26th floor of the destination floor (SA127) of the top row of FIG. 24 in addition to “75%” which is the probability (SA119) of occurrence of two or more of the top row of FIG. 23. Obtained by multiplying by "10%".
[0181]
　FIG. 25 shows an example, and when the destination floor prediction by time zone SP051 is used to express the destination floor forecast by time zone, if necessary data is added, the data is added by time zone. The destination floor prediction result SA13 can be changed.
[0182]
　FIG. 26 is an explanatory diagram of the rule/control template SA14 held by the analysis server SA according to the embodiment of this invention.
[0183]
　The rule/control template SA14 is a table that stores a template of elevator operation rules/control parameters. Here, the operation rule is a rule applied for the control panel CA to control the operation of a plurality of elevator cages that are subject to group management, and the control parameter is a parameter that can be changed in each operation rule. is there. In the present embodiment, the operation rule and the control parameters included therein are collectively referred to as an operation rule/control parameter. Further, the operation rule may be simply described as a rule, and the control parameter may be simply described as a parameter.
[0184]
　By using the rule/control template SA14, the optimum operation rule/control parameter can be searched. The search method consists of two steps. The first step is a search for rule/control No (SA140). This is a step of selecting a control parameter suitable for improving the KPI from a large number of operation rules/control parameters. The second step is a search for the parameter value (initial value) (SA144). The search target is a parameter value that can be controlled within the control parameter. By searching for this, more optimal control parameters can be obtained.
[0185]
　The rule/control No (SA140) is an ID for identifying an operation rule/control parameter. The rule name (SA141) is the name of the operation rule/control parameter. The condition (SA142) is an operation condition of the operation rule/control parameter. The parameter value (initial value) (SA143) is a controllable parameter in the operation rule/control parameter. For example, in the rule corresponding to the rule/control No. “Ru01”, “5 minutes later, direct flight from floor ◯”, the portion ◯ (floor number in this example) is a controllable parameter. The coefficient (initial coefficient) (SA145) is a coefficient for obtaining a regression equation or the like. The parameter value (initial value) (SA143) and the coefficient (initial coefficient) (SA145) can change the stored values ​​by repeating the optimization process.
[0186]
　The example shown in FIG. 26 is an example, and when the operation rule/control parameter of the elevator is realized, if there is necessary data, the rule/control template SA14 can be modified to add the data. ..
[0187]
　FIG. 27 is an explanatory diagram of the KPI list SA15 held by the analysis server SA according to the embodiment of this invention.
[0188]
　The KPI list SA15 is a table that stores a KPI (key performance indicator) that is an evaluation index when searching for an optimum operation rule/control parameter. Since the KPI may be different for each building, the KPI for each building is set in advance by the use flag (SA155). At that time, the target value of KPI (SA154) is also set.
[0189]
　The KPI ID (SA150) is an ID for identifying the KPI. The classification (SA151) is a classification of KPIs. Specifically, classification (SA151) shows who benefits by improving this KPI.
[0190]
　The name (SA152) is the name of the KPI. The condition (SA153) indicates the content of the KPI. The target value (SA154) indicates the target value of the changeable parameter value portion (circle in the example of FIG. 27) in the condition (SA153). Since this is different for each building, it is set before use. The usage flag (SA155) specifies a KPI to be used when implementing the optimization this time from a plurality of KIPs. When the use flag (SA155) is 1, it means to specify. Also, a plurality of KPIs may be designated.
[0191]
　In the example of FIG. 27, as the KPI, the waiting time until a person who appears at the landing gets into the basket, the congestion rate at the landing, and the amount of electricity used on the floor (that is, the amount of power consumption for moving the basket) are used. Shows. In these examples, for example, an operation rule/control parameter that reduces the maximum waiting time, an operation rule/control parameter that reduces the congestion rate of the landing, and an operation rule/control parameter that reduces the electricity usage are appropriate operations. Evaluated as a rule/control parameter.
[0192]
　However, the above is an example, and KPIs other than the above may be specified. For example, a KPI may be used such that the smaller the rate at which a plurality of people riding from different floors ride in the same basket, the higher the evaluation. Thereby, it is possible to realize the control of the cage in which the concerned person is less likely to be dissatisfied with the desire of the concerned person (for example, the user or the administrator) of the elevator.
[0193]
　The example shown in FIG. 27 is an example, and when the operation rule/control parameter of the elevator is realized, if there is necessary data, the KPI list SA15 can be changed to add the data.
[0194]
　FIG. 28 is an explanatory diagram of the simulation input and the result SA16 held by the analysis server SA according to the embodiment of this invention.
[0195]
　The simulation input and result SA16 is a table that stores the result processed by the KPI simulation processing SP11. The KPI simulation processing SP11 stores, as inputs, the generated person prediction result 2_SA11 indicating the occurrence situation, the destination floor prediction result SA13 by time zone, the rule/control template SA14 indicating the control parameter, and the KPI to be optimized. The KPI list SA15 is used. By using these data, it is possible to find operation rules/control parameters that increase the KPI when people occur.
[0196]
　In the KPI simulation process SP11, a process of outputting a KPI when a certain operation rule/control parameter is used is performed a plurality of times while changing the operation rule/control parameter in a state where people occur. The result is the simulation input and the result SA16.
[0197]
　The KPI simulation ID (SA160) is an ID for identifying the KPI simulation. The number of times (SA161) is the number of times when a plurality of KPI simulations were performed. The rule control list 1 (SA162) shows one set of operation rules/control parameters used in each simulation. The rule/control No (SA163) is an ID for identifying the operation rule/control parameter. The parameter value (SA164) is a control parameter used for this control. The coefficient (SA165) is a coefficient for obtaining a regression equation or the like. A plurality of rule control lists can be stored for one simulation. The KPI ID (SA166) is an ID for identifying the KPI. The KPI simulation result (SA167) is a value of KPI obtained as a result of KPI simulation using the rule control list.
[0198]
　The example shown in FIG. 28 is an example, and when the operation rules/control parameters of the elevator are realized, if there is necessary data, the simulation input and the result SA16 can be changed to add the data. it can.
[0199]
　FIG. 29 is an explanatory diagram of the effective rule/parameter SA17 held by the analysis server SA according to the embodiment of this invention.
[0200]
　The effective rule/parameter SA17 is a table for storing the result of obtaining the operation rule/control parameter contributing to the optimization (that is, effective) from the simulation input and the result SA16 shown in FIG. The rule/parameter evaluation unit SA38 receives the simulation input and the result SA16 shown in FIG. 28 as input, sets the objective variable as the KPI simulation result, uses the explanatory variable as the rule control list, and uses the results of multiple times to perform the multiple regression analysis. can do. However, it suffices that the rule control parameters contributing to the optimization can be specified, and therefore a method other than multiple regression analysis may be used.
[0201]
　The valid rule/parameter ID (SA170) is an ID for identifying a valid operation rule/control parameter. The effective rule control list 1 (SA171) is the rule control parameter that contributes the most to the multiple regression analysis. The rule/control No (SA172) is an ID that identifies an operation rule/control parameter. The parameter value (SA173) is the control parameter value used in this processing. The coefficient (SA174) is a coefficient obtained by multiple regression analysis and is a value indicating the degree of contribution to optimization. By referring to this, valid (that is, contributing to improvement of KPI) operation rule/control parameter is specified. A plurality of valid rule control lists can be stored. The KPI ID (SA175) is an ID for identifying the KPI. The predicted value (SA176) is the value of KPI predicted using the regression equation obtained by multiple regression analysis.
[0202]
　The example shown in FIG. 29 is an example, and when the operation rule/control parameter of the elevator is realized, if there is necessary data, the effective rule/parameter SA17 can be changed to add the data. ..
[0203]
　FIG. 30 is an explanatory diagram of the effective rule/parameter segmentation list SA18 held by the analysis server SA according to the embodiment of this invention.
[0204]
　Further optimization can be realized by subdividing the control parameter values ​​for the operation rules/control parameters identified from the effective rules/parameters SA17 shown in FIG. it can. An operation rule/control parameter having a large coefficient (SA174) is selected in the effective rule control list of the effective rule/parameter SA17. Then, the rule/parameter evaluation unit SA38 executes the effective rule/parameter subdivision processing SP14 for the selected operation rule/control parameter. Specifically, the rule/parameter evaluation unit SA38 can search for a more optimized operation rule/control parameter by increasing/decreasing the control parameter value included in the selected operation rule/control parameter.
[0205]
　The valid rule/parameter detailing ID (SA180) is an ID for identifying the valid rule/parameter detailing. The valid rule/parameter ID (SA181) is an ID for identifying a valid operation rule/control parameter. The effective rule control list 1 (SA182) is the rule control parameter estimated to contribute most by the multiple regression analysis. The rule/control No (SA183) is an ID that identifies an operation rule/control parameter. The parameter value (SA184) is the control parameter value used in this processing. The coefficient (SA185) is a coefficient obtained by multiple regression analysis and is a value that contributes to optimization. The parameter value subdivision range (SA186) is a value obtained by the effective rule/parameter subdivision processing SP14. A plurality of valid rule control lists can be stored. The KPI ID (SA187) is an ID for identifying the KPI. The predicted value (SA188) is the value of KPI predicted using the regression equation obtained by multiple regression analysis. Further, the rule/parameter evaluation unit SA38 may randomly select several operation rules/control parameters from the rule/control template SA14.
[0206]
　FIG. 30 is an example, and when realizing the elevator operation rule/control parameter, if there is necessary data, change the effective rule/parameter subdivision list SA18 so as to add the data. can do.
[0207]
　FIG. 31 is an explanatory diagram of the rule/parameter list SA19 held by the analysis server SA according to the embodiment of this invention.
[0208]
　The rule/parameter list SA19 is a table that stores a selection of operation rules/control parameters used in actual operation from the valid rules/parameters SA17 of FIG. Among the operation rules/control parameters in the effective rule control list, the one having a large coefficient (SA174) value is determined to be the operation rule/control parameter having a high contribution rate.
[0209]
　The rule/parameter ID (SA190) is an ID that identifies an operation rule/control parameter. The effective rule control first rank (SA191) is a rule control parameter estimated to have the largest degree of contribution as a result of multiple regression analysis. The rule/control No (SA192) is an ID that identifies an operation rule/control parameter. The parameter value (SA193) is the control parameter value used in this processing. The coefficient (SA194) is a coefficient obtained by multiple regression analysis and is a value that contributes to optimization.
[0210]
　The second effective rule control (SA195) is an operation rule/control parameter that is estimated to have the second largest contribution degree as a result of multiple regression analysis. The rule/control No (SA196) is the control parameter value used in this processing. The parameter value (SA197) is the control parameter value used in this processing. The coefficient (SA198) is a coefficient obtained by multiple regression analysis and is a value that contributes to optimization.
[0211]
　The KPI ID (SA199) is an ID for identifying the KPI. The predicted value (SA19A) is a value predicted using a regression equation obtained by multiple regression analysis.
[0212]
　The rule/parameter list SA19 is sent to the control selector SP06. The control selector SP06 generates an input command CA0 designating an operation rule/control parameter for improving the KPI based on the rule/parameter list SA19, and transmits it to the control panel CA. Based on the input command CA0, the control panel CA changes the already set operation rule/control parameter to the instructed one, and controls the car based on the changed operation rule/control parameter. This provides elevator control with improved KPI.
[0213]
　FIG. 31 is an example, and when implementing the operation rules/control parameters of the elevator, if there is necessary data, the rule/control parameter list SA19 may be modified to add the data. it can.
[0214]
　The processing described in the present embodiment is executed in the execution unit SA3 of the analysis server SA, but part or all of the processing may be executed by the control panel CA. For example, the control panel CA may have the same hardware as the analysis server SA shown in FIG. 1B, and at least part of the functions of the analysis server SA may be realized by these hardware.
[0215]
　FIG. 32 is an explanatory diagram of the building individualization report SA20 output by the analysis server SA according to the embodiment of this invention.
[0216]
　The building/personalization report SA20 is generated by the rule/parameter evaluation unit SA38 in the display/control data generation processing SP15, and is transmitted to the display unit SA1. The display unit SA1 (for example, a display device mounted as the output device 103) displays the received building individualization report SA20.
[0217]
　The building individualization report SA20 includes, for example, as shown in FIG. 32, a building name 3201, an elevator bank name 3202, a period 3203, a KPI 3204, and a result 3205.
[0218]
　The elevator bank name 3202 and the building name 3201 are the names of the elevator bank and the building in which it is installed, which is the target of the execution of each processing shown in FIG. 2, and the bank name (SA002) and the bank name (SA002) shown in FIG. Corresponds to the building name (SA007). A period 3203 is a period that is a simulation target. The KPI 3204 is an evaluation index selected as an evaluation target in the processing of the rule/parameter evaluation unit SA38, and corresponds to the KPI for which the usage flag (SA155) shown in FIG. 27 is valid. The result 3205 is a valid operation rule/control parameter selected as a result of the processing of the rule/parameter evaluation unit SA38, and corresponds to the operation rule/control parameter registered in the rule/parameter list SA19.
[0219]
　By referring to the building individualization report SA20, the elevator administrator can grasp the change content of the operation rule/control parameter necessary for improving the evaluation index displayed as the KPI3204. The administrator may manually set the grasped change of the operation rule/control parameter in the control panel CA. This provides elevator control with improved KPI.
[0220]
　As described above, according to the present embodiment, the number of people in the elevator hall is predicted from the number of people getting on and off the floor, a control method suitable for the prediction result is generated, and the control method uses an index related to dissatisfaction from the user. Optimal elevator control can be realized by performing evaluation. For example, by smoothly allocating the baskets to the landing near the time when future congestion is predicted, it is possible to suppress the long waiting time for the users at the landing, improve the transportation capacity of the user, and It is possible to improve the user's satisfaction level.
[0221]
　It should be noted that the present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail for better understanding of the present invention, and are not necessarily limited to those having all the configurations of the description.
[0222]
　Further, the above-described respective configurations, functions, processing units, processing means, etc. may be realized by hardware by designing a part or all of them, for example, by an integrated circuit. Further, each of the above-described configurations, functions, and the like may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as a program, a table, and a file for realizing each function is stored in a nonvolatile semiconductor memory, a hard disk drive, a storage device such as SSD (Solid State Drive), or a computer-readable non-readable memory such as an IC card, an SD card, or a DVD. It can be stored on a temporary data storage medium.
[0223]
　Further, the control lines and the information lines are shown as being considered necessary for explanation, and not all the control lines and the information lines in the product are necessarily shown. In reality, it may be considered that almost all the configurations are connected to each other.
claims
[Claim 1]
　An elevator analysis system having a processor and a storage device connected to the processor,
　wherein the storage device is the number of persons who appear to use an elevator at a landing on each floor of an elevator group to be controlled. Holds the number of occurrences, the
　processor,
　from the number of occurrences held in the storage device,
　predicts a future number of occurrences, from the predicted future number of occurrences, of each of the baskets belonging to the elevator group.
　An elevator analysis system characterized by determining an operation rule applied to control operation and control parameters set in each operation rule, and outputting the determined operation rule and control parameter.
[Claim 2]
　A elevator analysis system according to claim 1,
　wherein the storage device further holds a passenger number information indicating the actual ride number and off number at each floor of each car belonging to the elevator group,
　wherein the processor is
　the Based on the information on the number of people getting on and off, for each floor that can be the destination floor of the person who appeared at the landing of each floor, calculate the destination floor probability that is the probability that the destination floor of the person who appeared at the landing of each floor becomes that floor. ,
　An operation rule applied to control the operation of each of the baskets belonging to the elevator group, and a control parameter set in each of the operation rules is determined based on the future number of occurrences and the destination floor probability. An elevator analysis system characterized by the above.
[Claim 3]
　A elevator analysis system according to claim 2,
　wherein the storage device, the elevator group of the further holding information designating an evaluation index for evaluating the operation of each car,
　the processor,
　the future A plurality of first simulations in which a person is generated at a landing on each floor based on the number of occurrences and the destination floor probability to operate each of the baskets of the elevator group while changing the operation rule and the control parameter to be applied. The 　operation rule and the control parameter that are executed once
　,
calculate the specified evaluation index based on the result of the first simulation, and contribute to the improvement of the evaluation index based on the calculated evaluation index. Is determined as an operation rule applied to control the operation of each of the cars belonging to the elevator group, and a control parameter set in each operation rule.
[Claim 4]
　The elevator analysis system according to claim 3,
　wherein
　, based on the assumption that the probability distribution of the number of occurrences follows the Poisson distribution from the predicted number of occurrences in the future, the number of persons The
　elevator is characterized by calculating an occurrence probability that is the probability that the person appears, and causing the person to be generated according to the occurrence probability for each number of people and the destination floor probability for each destination floor and executing the first simulation. Analysis system.
[Claim 5]
　The elevator analysis system according to claim 3,
　further comprising a display device connected to the
　processor , wherein the processor has the operation rule that the degree of contribution to the improvement of the evaluation index satisfies a predetermined condition. And the said control parameter is specified, The
　said display device displays the specified said operation rule and control parameter, The elevator analysis system characterized by the above-mentioned.
[Claim 6]
　The elevator analysis system according to claim 3, further
　comprising: the processor, and an interface connected to a network outside the elevator analysis system,
　wherein the network controls each car belonging to the elevator group. The control device is connected, the
　processor
　identifies the operation rule and the control parameter in which the degree of contribution to the improvement of the evaluation index satisfies a predetermined condition
　, and the identified operation is performed via the interface. An elevator analysis system, which transmits rules and control parameters to the control device.
[Claim 7]
　The elevator analysis system according to claim 3,
　wherein the evaluation index is a waiting time until a person who rides occurs in any of the baskets, a congestion rate of the landing, and operating each basket of the elevator group. An elevator analysis system including any of the power consumption for
[Claim 8]
　The elevator analysis system according to claim 1,
　wherein the storage device includes
　information about the number of passengers getting on and off on each floor of each car belonging to the elevator group to be controlled and the number of people getting on and off the
　elevator group. Further holding operation log information indicating the actual state of each car to which it belongs, the
　processor
　generates a plurality of persons appearing at the landing of the elevator group in order to use the elevator, each of the landing of each person appears. The floor, the time when each person appears, and the destination floor of each person are randomly determined, and each car belonging to the elevator group is operated according to the time when each person appears, the floor of the landing where it appears, and the destination floor. By executing the second simulation, the number of generated persons, which is the number of persons appearing at the landing of each floor, is calculated from the state of each basket, the number of passengers of each basket on each floor, and the number of people of each basket getting off on each floor. By generating an estimated
　number of passengers estimation model and applying the actual number of passengers, the number of passengers getting off and the status of each basket obtained from the information on the number of passengers getting on and off and the operation log information, the number of persons generated on each floor Is estimated, and the
　estimated number of occurrences is held in the
　storage device, and the future number of occurrences is predicted from the estimated number of occurrences held in the storage device. ..
[Claim 9]
　The elevator analysis system according to claim 8
　, wherein, in the second simulation, the processor uses the number of persons on each floor for each predetermined time width as an objective index, and for each of the baskets for each predetermined time width. And the number of passengers of each car on each floor and the number of people of each car on each floor as an explanatory index, the elevator analysis system is characterized in that the generation number estimation model is generated.
[Claim 10]
　The elevator analysis system according to claim 8,
　wherein the boarding/alighting number information and the operation log information are stored in the elevator group when the number of passengers, the number of people getting off, and the actual state of each car are acquired. The information includes the operation rules that were applied to control the operation of each of the baskets that belong and information indicating the control parameters that were set in the operation rules, and the
　processor is the applied operation rules and the settings. The elevator analysis system, wherein the second simulation is executed by operating each of the baskets according to the control parameters.
[Claim 11]
　An elevator analysis method executed by an elevator analysis system having a processor and a storage device connected to the processor,
　wherein the storage device uses an elevator at a landing of each floor of an elevator group to be controlled. The number of persons that has appeared for the number of occurrences is held, and in the
　elevator analysis method, the
　processor predicts a future number of occurrences from the number of occurrences held in the storage device, and the
　processor. Is a second procedure for determining an operation rule applied to control the operation of each car belonging to the elevator group and a control parameter set in each operation rule from the predicted number of future occurrences. And a
　third procedure in which the processor outputs the determined operation rule and control parameter, the elevator analysis method.
[Claim 12]
　The elevator analysis method according to claim 11,
　wherein the storage device further retains boarding/alighting passenger information indicating the actual number of passengers and the number of passengers on each floor of each car belonging to the elevator group, and in
　the second procedure. The processor is a
　probability that the destination floor of the person who appears at the landing of each floor becomes the floor for each floor that can be the destination floor of the person who appears at the landing of each floor based on the information on the number of people getting on and off. Calculate the
　destination floor probability, based on the future number of people and the destination floor probability, the operation rule applied to control the operation of each of the baskets belonging to the elevator group, and set in each operation rule An elevator analysis method, characterized in that a control parameter to be determined is determined.
[Claim 13]
　The elevator analysis method according to claim 12,
　wherein the storage device further holds information designating an evaluation index for evaluating the operation of each of the baskets of the elevator group, and in
　the second procedure, The processor
　applies the first simulation in which a person is generated at a landing of each floor based on the future number of occurrences and the destination floor probability and operates each of the cages of the elevator group, and the operation rule and the control are applied. Executed a plurality of times while changing parameters
　,
　calculating the designated evaluation index based on the result of the first simulation, and contributing to improvement of the evaluation index based on the calculated evaluation index. An elevator analysis characterized in that the operation rule and the control parameter are determined as an operation rule applied to control the operation of each car belonging to the elevator group, and a control parameter set in each operation rule. Method.
[Claim 14]
　The elevator analysis method according to claim 11,
　wherein the storage device stores the
　number of passengers in the
　elevator group, which indicates the actual number of passengers and the number of passengers on each floor of each car belonging to the elevator group to be controlled, and the elevator group. Further holding operation log information indicating the actual state of each car belonging, the
　elevator analysis method, further, the
　processor, to generate a plurality of persons appearing at the landing of the elevator group to use the elevator. , Randomly determining the floor of the landing where each person appears, the time when each person appears, and the destination floor of each person, and the elevator according to the time when each person appears, the floor of the landing that appears, and the destination floor. By executing the second simulation of operating each car belonging to the group, the appearance of the car at each floor appeared from the state of each car, the number of passengers of each car on each floor, and the number of people of each car getting off on each floor. A procedure for generating an occurrence number estimation model that estimates the number of persons that is the number of persons, and the
　processor, the actual number of passengers obtained from the boarding/alighting number information and the operation log information, the number of people getting off, and the state of each basket by applying the generated number estimation model, the procedure of estimating the floor of the generation number,
　the processor, the generation number of people estimated comprises, a step of holding in the storage device,
　in the first procedure, wherein the processor Is a method of analyzing an elevator, wherein the future number of occurrences is predicted from the estimated number of occurrences held in the storage device.