Description:FIELD OF THE INVENTION
The present invention relates to the technical field of electric vehicles. In particular, the present invention relates to a system for predicting a remaining driving range of the electric vehicle and a method thereof.
BACKGROUND OF THE INVENTION
Electric vehicles have gained popularity due to their eco-friendly nature and potential to reduce dependence on fossil fuel. However, range anxiety is still one of the prominent reasons customers hesitate to switch to an electric vehicle. For the range algorithm to be as accurate as possible, it needs to take the wide range of rider’s driving patterns into consideration. Range estimation has been implemented in various configurations in electric vehicles. A very prominent method, for example, is the Wh/km efficiency figure, where the battery capacity’s usage with the vehicle’s mileage is compared, and based on that efficiency, the further range is calculated using the remaining available battery capacity. Likewise, another method is to calculate the SoC drop rate of the vehicle and then estimate the further range based on the SoC remaining figure.
A patent application WO2020211456A1 titled ”A method of measuring remaining range of electric vehicle, electronic device, and storage medium” discloses a method comprising acquiring historical traveling data of an electric vehicle and a historical remaining range corresponding to the historical traveling data, wherein the traveling data at least comprises the remaining battery power of the electric vehicle, using the historical traveling data as an input object and the historical remaining range as a monitored object, and creating and training a prediction model, wherein current traveling data serves as an input to the prediction model, and an estimated remaining range serves as an output from the prediction model, acquiring current traveling data of the electric vehicle, inputting the current traveling data into the trained prediction model, and obtaining, from the prediction model, an estimated remaining range regarding the current traveling data and displaying the estimated remaining range.
Another patent application CN111670340B titled “Method for acquiring remaining driving mileage of vehicle, electronic equipment and vehicle” discloses a method for acquiring the remaining driving mileage of the vehicle comprising the steps: acquiring a current position and a current remaining energy of a vehicle; acquiring user historical driving data; the remaining driving range of the vehicle is determined according to the user historical driving data, the current position of the vehicle and the current remaining energy.
Yet another patent application CN110077274B titled “Estimation method, device and equipment for travelling distance of logistics electric vehicle” discloses a method comprising the following steps: acquiring a departure point, a destination and environmental information of a driving path between the departure point and the destination; acquiring the current load capacity and battery information of the logistics electric vehicle; and determining the actual driving distance of the logistics electric vehicle in the driving process based on the driving path according to the current load capacity, the battery information and the environment information.
The primary issue with the above methods is that none of the above methods are tailored to the user’s driving pattern. Further, none of the above stated prior arts take historic data into account during prediction. In addition, current drawn by the motor, real-time motor speed and load of the vehicle are not taken into consideration together for such calculations. An efficient and accurate range estimation algorithm is a must for any electric or hybrid vehicle. Existing methods for predicting the remaining driving range of electric vehicles often suffer from limitations and inaccuracies, leading to inconveniences and anxiety about running out of power in inappropriate situations.
In order to overcome the aforementioned drawbacks, there is a need to provide a system and a method for predicting a remaining driving range of an electric vehicle by taking into consideration data values of historical parameters from historical trips and data values of current parameters for the ongoing trips, and calculating the remaining range using these parameters.
OBJECTIVES OF THE DISCLOSURE
A primary objective of the present invention is to overcome the disadvantages of the prior-arts.
Another objective of the present disclosure is to predict the estimated range of a vehicle using the trained range prediction model having a set of input variables and an output variable.
Yet another objective of the present disclosure is to implement a driving pattern-based approach to provide a range estimation to the driver of the vehicle.
SUMMARY OF THE INVENTION
The following is a summary description of illustrative embodiments of the invention. It is provided as a preface to assist those skilled in the art to more rapidly assimilate the detailed design discussion which ensues and is not intended in any way to limit the scope of the claims which are appended hereto in order to particularly point out the invention.
An embodiment of the present invention relates to a method for predicting a remaining driving range for a vehicle. The method comprising the steps of determining, by one or more processors, a plurality of historical data values for a first set of historical driving data variables and a historical driving range corresponding to the first set of historical driving data variables for a plurality of past trips, wherein the plurality of historical data values of the first set of historical driving data variables are collected over multiple time intervals; grouping, by the one or more processors, the plurality of historical data values of each of the historical driving data variables of the first set of historical driving data variables into a corresponding plurality of bins to form a histogram of each of the historical driving data variables of the first set of historical driving data variables. A historical data value for each of the historical driving data variables of a second set of historical driving data variables are also determined by the one or more processor.
Further, the method comprising training based on the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range, a range prediction model, wherein based on the training, the range prediction model learns a correlation between the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range. Furthermore, the method comprises, determining a plurality of current data values for a first set of current driving data variables for the ongoing trip, wherein the plurality of current data values of the first set of current driving data variables are collected over multiple time intervals. Also, the method comprising, grouping of the plurality of current data values of each of the current driving data variables of the first set of current driving data variables into a corresponding plurality of bins to form a histogram of each of the current driving data variable of the first set of current driving data variables by one or more processor, an instantaneous data value for each of the current driving data variables of a second set of current driving data variables for the ongoing trip are also determined for the prediction of remaining range of the vehicle.
The final step of the method includes predicting the remaining driving range of the vehicle using the trained range prediction model based on the second set of current driving data variables, and at least one of the histograms of the first set of current driving data variables and the first set of historical driving data variables.
The method further comprises collecting the current data values of the first set of the current driving data variables, and subsequently modifying the corresponding existing historical data values of the first set of historical driving data variables with the current data values of the first set of current driving data variables of the ongoing trip.
In accordance with an embodiment of the present invention, the first set and the second set of historical and current driving data variables include at least one of a battery state-of-charge (SoC), a vehicle load, a motor speed, and a current drawn by the motor.
In accordance with an embodiment of the present invention, during the training step of the method, the first and second set of historical driving data variables are used as input variables, and the historical driving range is used as an output variable.
In accordance with an embodiment of the present invention, the multiple time intervals include, but are not limited to, 1 minute, 10 minutes, 1 hour, and 10 hours.
In accordance with an embodiment of the present invention, the format of the data values for the historical driving data variables and the current driving data variables is one of a real value, an absolute value, and a percentage value.
In accordance with an embodiment of the present invention, the plurality of historical data values of the first set of historical driving data variables are pre-stored in the vehicle during the vehicle production phase.
In accordance with an embodiment of the present invention, the input to the trained range prediction model includes, instantaneous values of the second set of current driving data variables, and a one-minute average of at least one of the histograms of the first set of historical driving data variables and the first set of current driving data variables.
In accordance with an embodiment of the present invention, the one-minute average is calculated using one of a moving average, a running average, and an average in general based on the computational ability. Further, the one-minute average of the first set of historical driving data variables is calculated using the plurality of historical data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving. Likewise, the one-minute average of the first set of current driving data variables is calculated using the plurality of current data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving.
In accordance with an embodiment of the present invention, the vehicle range prediction corresponds to a plurality of driving modes available in the vehicle which includes, but not limited to, an eco-mode, a ride mode, a rush mode, a sport mode, and a limp mode.
In accordance with an embodiment of the present invention, predicting the vehicle range corresponds to a plurality of driver profiles or user profiles available in the vehicle. The driver profile or user profile correspond to a driving pattern or a driving behaviour of a specific user.
In accordance with an embodiment of the present invention, the method further captures the driving data values for a first set of current driving data variables for the ongoing trip based on the selected driver profile and the driving mode for the selected driver profile.
In accordance with an embodiment of the present invention, the trained range prediction model is configured either locally on-board the vehicle, or on a remote computing platform.
In accordance with an embodiment of the present invention, the range prediction model is trained corresponding to at least one of a plurality of driving modes available in the vehicle, at least one of a plurality of driver profiles available, and a combination thereof.
In accordance with an embodiment of the present invention, the method of predicting the remaining driving range for the vehicle is performed by one or more control units configured with one or more processors. In particular, the control unit of the vehicle includes, but not limited to, one of a vehicle control unit (VCU), an electronic control unit (ECU) and a motor control unit (MCU).
In accordance with an embodiment of the present invention, the method may be performed by a separate one or more processing units.
In accordance with an embodiment of the present invention, the method further comprises the step of displaying the predicted remaining range of the vehicle to a user or a rider. In particular, the predicted remaining range of the vehicle is displayed to the user on a display device that is touch-sensitive or non-touch sensitive, and including, but not limited to, one of a CRT display, an LCD display, an LED display, an OLED display, an AMOLED display, and a PMOLED display. In an embodiment, the display device is a vehicle display or the display of a user’s handheld device.
Further, in various embodiments, the vehicle is one of, but not limited to, a battery electric vehicle (BEV), a hybrid electric vehicle (HEV), a Plug-in Hybrid electric vehicle (PHEV), a Fuel Cell electric vehicle (FCEV). Further, the vehicle is a two-wheeled vehicle or a multi-wheeled vehicle.
BRIEF DESCRIPTION OF THE DRAWINGS
To describe the technical solutions in the embodiments of the present disclosure or in the prior art more clearly, the following briefly describes the accompanying drawings required for describing the embodiments or the prior art. Apparently, the accompanying drawings in the following description merely show some embodiments of the present disclosure, and a person of ordinary skill in the art can derive other implementations from these accompanying drawings without creative efforts. All of the embodiments or the implementations shall fall within the protection scope of the present disclosure. Having thus described the disclosure in general terms, reference will now be made to the accompanying figures.
Fig. 1 is a block diagram illustrating a range prediction system for use within a vehicle environment to predict a remaining driving range of the vehicle in accordance with an embodiment of the invention;
Fig. 2 is a block diagram illustrating different control units, in accordance with an embodiment of the present disclosure.
Fig. 3 is a flow chart illustrating a method to predict remaining driving range of the vehicle in accordance with an embodiment of the present invention;
Fig. 4 is a flow chart illustrating a method to predict remaining driving range of an electric vehicle in accordance with an exemplary embodiment of the present invention.
It should be noted that the accompanying figure is intended to present illustrations of a few examples of the present disclosure. The figure is not intended to limit the scope of the present disclosure. It should also be noted that the accompanying figure is not necessarily drawn to scale.
DETAILED DESCRIPTION OF THE INVENTION
In the following detailed description of the invention, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be obvious to a person skilled in the art that the invention may be practiced with or without these specific details. In other instances, well known methods, procedures and components have not been described in detail so as not to unnecessarily obscure aspects of the invention.
The accompanying drawing is used to help easily understand various technical features and it should be understood that the alternatives presented herein are not limited by the accompanying drawing. As such, the present disclosure should be construed to extend to any alterations, equivalents and substitutes in addition to those which are particularly set out in the accompanying drawing. Although the terms first, second, etc. may be used herein to describe various elements or values, these elements or values should not be limited by these terms. These terms are generally only used to distinguish one element or values from another.
It will be apparent to those skilled in the art that the other alternatives of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention. While the foregoing written description of the invention enables one of ordinary skill to make and use what is considered presently to be the best mode thereof, those of ordinary skill will understand and appreciate the existence of variations, combinations, and equivalents of the specific aspect, method, and examples herein. The invention should therefore not be limited by the above described alternative, method, and examples, but by all aspects and methods within the scope of the invention. It is intended that the specification and examples be considered as exemplary, with the true scope of the invention being indicated by the claims.
Conditional language used herein, such as, among others, "can," "may," "might," "may," “e.g.,” and the like, unless specifically stated otherwise, or otherwise understood within the context as used, is generally intended to convey that certain alternatives include, while other alternatives do not include, certain features, elements and/or steps. Thus, such conditional language is not generally intended to imply that features, elements and/or steps are in any way required for one or more alternatives or that one or more alternatives necessarily include logic for deciding, with or without other input or prompting, whether these features, elements and/or steps are included or are to be performed in any particular alternative. The terms “comprising,” “including,” “having,” and the like are synonymous and are used inclusively, in an open-ended fashion, and do not exclude additional elements, features, acts, operations, and so forth. Also, the term “or” is used in its inclusive sense (and not in its exclusive sense) so that when used, for example, to connect a list of elements, the term “or” means one, some, or all of the elements in the list.
Terms vehicle and electric vehicle may interchangeably be used for convenience.
Terms vehicle control unit or VCU may interchangeably be used for convenience.
Terms motor control unit or MCU may interchangeably be used for convenience.
Terms electronic control unit or ECU may interchangeably be used for convenience.
Terms battery management system or BMS may interchangeably be used for convenience.
Terms range prediction model/system or trained prediction model/system may interchangeably be used for convenience.
Terms historical driving parameters or historical driving variables or historical driving data variables may interchangeably be used for convenience.
Terms current driving parameters or current driving variables or current driving data variables may interchangeably be used for convenience.
Terms data value or data points may interchangeably be used for convenience.
Terms input parameters or input data variables may interchangeably be used for convenience.
Fig. 1 is a block diagram of a range prediction system 100 for use within a vehicle environment to predict remaining driving range of the vehicle in accordance with one embodiment of the invention. The range prediction system 100 resides locally on-board the vehicle for predicting the remaining range of the vehicle.
The range prediction system 100 operating in a vehicle environment includes a battery management system 102, one or more control units 104, and a display device 110. In particular, the range prediction system 100 predicts the range of the vehicle by first training a range prediction model. The range prediction model is trained using a first and a second set of historical driving data variables that are used as input variables, and the historical driving range is used as an output variable. More specifically, the range prediction model is trained using the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range for a plurality of past trips. The trained range prediction model of the range prediction system 100 develops a correlation between the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range.
In accordance with an embodiment of the present invention, the vehicle is one of an electric vehicle, a hybrid electric vehicle, a fuel-cell electric vehicle. Further, the vehicle is a two-wheeled vehicle or a multi-wheeled vehicle. The vehicle is not limited to a battery electric vehicle (BEV), a hybrid electric vehicle (HEV), a Plug-in Hybrid electric vehicle (PHEV), Fuel Cell electric vehicle (FCEV), a two wheeler electric bike, and a three wheeler electric vehicle.
In accordance with an embodiment of the present invention, the one or more control units 104 is configured with one or more processors 106 and a memory 108. The one or more processors 106 is in communication with the memory 108 to perform a series of computer-executable instructions stored in the memory such as, determining a plurality of historical data values for the first set of historical driving data variables and the historical driving range corresponding to the first set of historical driving data variables for a plurality of past trips; grouping the plurality of historical data values of each of the historical driving data variables of the first set of historical driving data variables into a corresponding plurality of bins to form a histogram of each of the historical driving data variables of the first set of historical driving data variables; determining a historical data value for each of the historical driving data variables of a second set of historical driving data variables, and training a range prediction model based on the first set of historical driving data variables, the second set of historical driving data variables and the historical driving range. Further, determining and grouping a plurality of current data values for a first set of current driving data variables for the ongoing trip; determining an instantaneous data value for each of the current driving data variables of a second set of current driving data variables for the ongoing trip, and lastly, predicting the remaining driving range of the vehicle using the trained prediction model.
The one or more processors 106 associated with the one or more control units (CU) 104 may be any well-known processor, but not limited to processors from Intel Corporation. Alternatively, the processor may be a dedicated controller such as an ASIC or ARM, MIPS, SPARC, or INTEL® IA-32 microcontroller or the like. The one or more control units 104 is anyone of the vehicle control unit (VCU) 202, an electronic control unit and the motor control unit (MCU) 204.
In yet another embodiment of the present invention, the one or more processors 106 comprise a collection of processors which may or may not operate in parallel. Alternatively, the one or more processors 106, which may be any processor-driven device, such as one or more microprocessors and memories or other computer-readable media operable for storing and executing computer-readable instructions.
The memory 108 associated with the one or more control units (CU) 104 stores instructions to be executed by the one or more processors 106. The memory 108 can be any type of suitable memory, including various types of dynamic random access memory (DRAM) such as SDRAM, various types of static RAM (SRAM), and various types of non-volatile memory (PROM, EPROM, and flash). It should be understood that the memory 108 may be a single type of memory component or it may be composed of many different types of memory components. As noted above, the memory 108 stores instructions for executing one or more methods for estimating remaining driving range of the vehicle. For example, the memory 108 may store software used by the user device, such as an operating system (not shown), application programs (not shown), and an associated internal database (not shown). In an alternative embodiment, the memory 108 may be an external memory to the one or more control unit (CU) 104 to store various data and computer executable instructions.
The one or more control units (CU) 104 include one or more automotive control units for controlling the various on-board systems of the vehicle. In particular, the one or more control units (CU) 104 includes at least one of, but is not limited to, a vehicle control unit (VCU) 202, and a motor control unit (MCU) 204. The one or more control units 104 is further explained with respect to Fig. 2 which is described in detail below. Each control unit (CU) 104 of the one or more control units 104 may include the memory 108, and the one or more processors 106 including other components that control the operation of the vehicle. The one or more control units 104 are implemented locally or on-board on the vehicle. The one or more processors 106 executes different instructions to perform the method for predicting the remaining driving range of the vehicle.
In an alternative embodiment, the one or more control units (CU) 104 may not be fully implemented locally or on-board on the vehicle. The one or more control units (CU) 104 may be configured with a computing device, such as a remote server or a cloud.
In an alternate embodiment of the present invention, a separate processing unit may perform all the instructions to predict the remaining range of the vehicle.
In an embodiment of the present invention, the one or more processors 106, collect a plurality of historical data values for the first set of historical driving data variables and a historical driving range corresponding to the first set of historical driving data variables for a plurality of past trips, wherein the plurality of historical data values of the first set of historical driving data variables are collected over multiple time intervals. Further, the one or more processors 106, perform grouping of the plurality of historical data values of each of the historical driving data variables of the first set of historical driving data variables into a corresponding plurality of bins to form the histogram of each of the historical driving data variables of the first set of historical driving data variables. The one or more processors 106 also determine the historical data value for each of the historical driving data variables of the second set of historical driving data variables. Further, the range prediction model is trained based on the 1-minute average of the histogram of the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range. The training of the range prediction model helps learn a correlation between the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range.
The one or more processors 106 perform similar collection of the plurality of current data values for the first set of current driving data variables for the ongoing trip over multiple time intervals, and grouping of the plurality of current data values of each of the current driving data variables of the first set of current driving data variables into the corresponding plurality of bins to form a histogram of each of the current driving data variable of the first set of current driving data variables. Further, instantaneous data value for each of the current driving data variables of the second set of current driving data variables for the ongoing trip is also determined.
Further, the trained range prediction model predicts the remaining driving range of the vehicle, based on the second set of current driving data variables, and at least one-minute average of one of the histograms of the first set of current driving data variables and the first set of historical driving data variables. In accordance with an embodiment of the present invention, the first set and the second set of historical and current driving data variables include at least one of a battery state-of-charge (SoC), a vehicle load, a motor speed, and a current drawn by the motor.
In accordance with an embodiment of the present invention, the battery management system (BMS) 102 is dedicated to the oversight of a battery pack of the vehicle, which is an assembly of battery cells. The battery management system (BMS) 102 provides battery state-of-charge (SoC) in real time or near real-time to the control unit 104. Also, the values related to current drawn by the motor is provided by the battery management system (BMS) 102 to the control unit 104. The battery management system (BMS) 102 provides both these data values i.e., SoC and current drawn by the motor to the control unit 104, such as vehicle control unit (VCU) 202 to estimate the remaining range of the vehicle.
In accordance with an exemplary embodiment of the present disclosure, the motor speed and the current drawn by the motor of the vehicle are collected over multiple time intervals and distributed into the plurality of bins to form the histogram, respectively. The individual bins of the plurality of bins in the corresponding histograms of motor speed and current drawn by the motor include the different range of values for motor speed and current drawn by the motor, respectively.
For example, in the histogram of motor speed, the lower values of motor speed are grouped into lower order bins, and the higher values of motor speed are grouped into higher order bins, wherein each of the bins of the plurality of bins are of equal width. Similarly, in the histogram of current drawn by the motor, the lower values of current drawn by the motor are grouped into lower order bins, and the higher values of current drawn by the motor are grouped into higher order bins, wherein each of the bins of the plurality of bins are of equal width.
In accordance with an embodiment of the present invention, the histogram includes data points collected over multiple time intervals. The multiple time intervals include, but may not be limited to, 1 minute, 10 minute, 1 hour and 10 hours. In an example, the histogram of motor speed is built by grouping the different motor speed data points or data values of the vehicle collected over multiple time intervals into 10 groups or 10 bins. For example, bin 1 includes all the data points of motor speed ranging from 0 to 10 rpm, bin 2 includes all the data points of motor speed ranging from 10 to 20 rpm, so on and so forth until bin 10 which includes all the data points of motor speed ranging from 90 to 100 rpm. The different motor speed data points are collected in every one minute considering the sampling time of 0.1 second, for example. Therefore, 600 different motor speed data points would be collected in one minute. Likewise, 6000 motor speed data points in 10 minutes, 36,000 motor speed data points in 1 hour and 360,000 motor speed data points in 10 hours would be collected. The number of data points collected for a driving parameter (for e.g., motor speed in this case) within a particular time window, such as 1 minute, 10 minutes, 1 hour or 10 hours can be increased or decreased depending upon the chosen sampling time for recording a data point. Further, the number of bins in the histogram may be any other suitable number and does not necessarily always include 10 bins. In an alternative embodiment, the rpm value of the motor speed may directly be converted into its corresponding percentage value, and accordingly, grouped into a relevant bin of the motor speed histogram. For example, bin 1 includes all the data points of motor speed ranging from 0 to 10%, bin 2 includes all the data points of motor speed ranging from 10 to 20%, so on and so forth until bin 10 which includes all the data points of motor speed ranging from 90 to 100%. Likewise, the data points for the input parameter, current drawn by the motor, would be collected and then these data points would be grouped into its relevant bin to form the histogram for the current drawn by the motor.
In an implementation, the current data values of the first set of the current driving data variables are collected for the ongoing trip, and subsequently modifying the corresponding existing historical data values of the first set of historical driving data variables with the current data values of the first set of current driving data variables of the ongoing trip in Last in First Out (LIFO) fashion, when moving average is used. In other words, in the case of moving average, the historical or older data values move out of the range prediction system or get replaced with the new or current data values captured from the ongoing trip. Further, in case of running average, the historical or older data values of the first set of historical driving data variables will not move out of the range prediction system although the weightage for the historical or older data values will keep on reducing as new and current driving data values of the first set of current driving data variables for the on-going trip will keep on recording.
In another implementation, the one-minute average of the current and historical driving data variables may be calculated by anyone of the running average, the moving average or an average in general based on the computational ability.
In another exemplary embodiment of the present invention, the histogram is independently built for motor speed and current drawn by the motor. The one-minute average for each of these histograms may be calculated using one of a moving average and a running average. For example, in a case when moving average is implemented, the prestored historical data values in the histograms of motor speed and current drawn by the motor gets replaced or overwritten with currently captured data values of motor speed and current drawn by the motor in every 1 minute of the ongoing trip in the Last in First Out (LIFO) order. The one-minute average for the histogram of motor speed and the histogram of current drawn by the motor is calculated using the corresponding data values recorded in 1 minute, 10 minutes, 1 hour and 10 hours of driving window. Further, the window size for collecting data values i.e., 1 minute, 10 minutes, 1 hour and 10 hours is dynamic and may be decided based on one’s requirements. In another example, in a case when running average is implemented, the prestored historical data values in the histograms of motor speed and current drawn by the motor do not get replaced or overwritten with currently captured data values of motor speed and current drawn by the motor. Instead, the historical data values as well as the currently captured data values for the ongoing trip both are collected and retained. However, the weightage is proportionally reduced for the data values of the older data points as new data points from the current ongoing trip keeps on accumulating. For example, for the ongoing trip, if 1 minute of current driving data values for motor speed and current drawn by the motor is captured then the prestored historical data values of motor speed and current drawn by the motor for the last 10 minutes, 1 hour and 10 hours is proportionally updated. In other words, highest weightage is assigned to the data values of the data points which are currently captured in 1 minute of the ongoing trip and less weightage is assigned to the rest of the data values for the older or historical data points.
The histograms of motor speed and current drawn by the motor are either stored locally on-board the vehicle, or on the server or the cloud computing platform.
In an exemplary embodiment of the present invention, the histogram of motor speed and the histogram of current drawn by the motor which includes the historical data values collected over multiple time intervals are pre stored in the vehicle during the vehicle production phase. Further, range prediction system 100 via one or more processors 106 collects the plurality of current data values for the first set of current driving data variables for the ongoing trip, and subsequently modifies or replaces the corresponding pre stored existing data values in the histograms of the first set of historical data variables with the plurality of current data values of the first set of current driving data variables.
In accordance with an embodiment of the present invention, the input to the trained range prediction model includes, instantaneous value of the remaining battery state-of-charge (SoC) as measured by the BMS in real-time or near real-time, vehicle mass/load as measured by the control unit 104 in real time or near real-time, and one-minute average of at least one of the histograms of the first set of historical driving data variables and the first set of current driving data variables. The prediction model calculates and provides the remaining range of the vehicle in accordance with the above mentioned parameters.
In accordance with an embodiment of the present invention, a weight is assigned to each of the input parameters as a result of training the range prediction model. In another alternative embodiment, separate weights may be assigned corresponding to each of the bins in the histograms of motor speed and current drawn by the motor, respectively. A trained prediction model including a separate weight for each of the input parameters is generated. The trained range prediction model provided locally on-board the vehicle starts capturing the driving pattern of the vehicle when the user starts driving the vehicle. In general, the driving pattern refers to the data values of the input parameters, such as motor speed and the current drawn by the motor. When the user starts driving the vehicle, to predict the vehicle range, the input to the trained prediction model includes the instantaneous value of battery state of charge as measured by the BMS, the vehicle load as estimated by the control unit 104 instantaneously, and one-minute average calculated for the input parameters, such as for the histograms of the motor speed and the current drawn by the motor.
In an exemplary embodiment of the present invention, the prediction model displays the estimated remaining range of the vehicle on the display device 110. The current motor speed and the corresponding current drawn by the motor overwrites the pre-stored data values in the histograms of the motor speed and the current drawn by the motor in Last in First Out (LIFO) order as described above. In an example, if the user drives the vehicle only for 1 minute, then 1 minute of current data i.e., 1 minute of captured current data values for motor speed and current drawn by the motor replaces or overwrites the existing 1 minute of historical data values of motor speed and the current drawn by the motor in Last in First Out (LIFO) order. The one-minute average of motor speed for example, is then calculated using the currently captured 1-minute current data values of motor speed, and the historical data values of motor speed for pre-stored past 9 minutes, 59 minutes and 9 hours 59 minutes. Likewise, the one-minute average for current drawn by the motor is calculated. Further, when the user has driven the vehicle for 10 minutes, the 10 minutes of currently captured data values for motor speed and current drawn by the motor replaces or overwrites the existing 10 minutes of historical data values of motor speed and current drawn by the motor in Last in First Out (LIFO) order, and then the one-minute average is calculated for motor speed and current drawn by the motor using the combination of currently captured data value of 1 minute and 10 minutes, and historical data values of motor speed and current drawn by the motor for pre-stored past 50 minutes and 9 hours and 50 minutes. Further, the pre-stored data values of motor speed and current drawn by the motor is completely overwritten with the current driving data values when the driver has driven the vehicle for 10 hours or the maximum time interval used for grouping of the first set of current driving data variables. Furthermore, the pre-stored data values of motor speed and current drawn by the motor keep on over-writing with the new or current data values of motor speed and current drawn by the motor for the ongoing trip while driving.
In an embodiment of the present invention, the trained range prediction model assigns weights to each of the plurality of historical driving data variables corresponding to each of the plurality of driving modes available in the vehicle. In another embodiment, there may be a separate range prediction model trained corresponding to each of the plurality of driving modes available in the vehicle. The plurality of driving modes includes, but not limited to, one of an eco-mode, a ride mode, a rush mode, a sport mode, and a limp mode. In another embodiment, the weights assigned to each of the input parameters of the corresponding trained prediction models may be different corresponding to the vehicle drive modes. For example, the weights of the input parameters of the range prediction model for eco mode may be different from the weights of the input parameters of the range prediction model for the ride mode. Therefore, when a driver of the vehicle selects a particular driving mode, the range prediction model corresponding to the selected drive mode is activated for predicting the vehicle range.
In an embodiment of the present invention, the trained range prediction model assigns weights to each of the plurality of historical driving data variables corresponding to each of the plurality of driver or user profiles available in the vehicle. In another embodiment, there may be a separate range prediction model trained corresponding to each of the plurality of driver or user profiles available in the vehicle. In another embodiment, the weights assigned to each of the input parameters of the corresponding trained prediction models may be different corresponding to the driver or user profiles. For example, the weights of the input parameters of the range prediction model for driver profile 1 may be different from the weights of the input parameters of the range prediction model for the driver profile 2. Therefore, when a driver of the vehicle selects a particular driving profile, the range prediction model corresponding to the selected driver profile is activated for predicting the vehicle range.
In an alternative embodiment of the present invention, the range prediction model may be trained corresponding to at least one of a plurality of driving modes available in the vehicle, at least one of a plurality of driver profiles available in the vehicle, and a combination thereof. For example, the weights of the input parameters of the range prediction model for driver profile 1 selecting an eco-mode for driving may be different from the weights of the input parameters of the range prediction model for the driver profile 1 selecting a rush mode for driving. In another example, the weights of the input parameters of the range prediction model for driver profile 1 selecting a sport mode for driving may be different from the weights of the input parameters of the range prediction model for the driver profile 2 selecting a sport mode for driving.
In an example, the equation for range measurement by the prediction model is as follows:
Y =SoC(w_1⋅Speed Histogram)+ SoC(w_2⋅Current Histogram)+ (w_3⋅Vehicle Load)+ Range Bias
where, w1, w2, w3 are the weights assigned to the input variables while training a prediction model, and Y is the vehicle range to be measured.
In accordance with an embodiment of the present invention, the display device 110 displays the predicted remaining range of the vehicle to the driver. The display device 110 is attached to the vehicle and visible to the driver while driving the vehicle and it includes a touch-sensitive display device or non-touch sensitive display device. The display device 110 also includes, but is not limited to, one of a CRT display, an LCD display, an LED display, an OLED display, an AMOLED display, and a PMOLED display.
In an alternate embodiment of the present invention, the display device 110 may communicatively be coupled to a user handheld device 112 to display the predicted range of the vehicle. The user handheld devices 112 include one of a desktop computer, a laptop computer, a user computer, a tablet computer, a personal digital assistant (PDA), a cellular telephone, a communication network appliance, a camera, a smartphone, an enhanced general packet radio service (EGPRS) mobile phone, a media player, a navigation device, an email device, a game console, or a combination of any these data processing devices or other data processing devices. Furthermore, the user handheld devices 112 may be any user handheld device that can be provided access to and/or receive application software executed and/or stored on any of the servers.
In some implementations, the user handheld devices 112 can communicate wirelessly to the display device 110 of the vehicle through a communication network (not shown in the figure) such as, but not limited to, the Internet, wireless networks, local area networks, wide area networks, private networks, a cellular communication network, corporate network having one or more wireless access points or a combination thereof connecting any number of mobile clients, fixed clients, and servers and so forth. Examples of the communication network 108 may include the Internet, a WIFI connection, a Bluetooth connection, a Zigbee connection, a communication network, a wireless communication network, a 3G communication, network, a 4G communication network, a 5G communication network, a USB connection, or any combination thereof. For example, the communication may be based through a radio-frequency transceiver (not shown). In addition, short-range communication may occur, such as using Bluetooth, Wi-Fi, or other such transceivers.
Fig. 2 is a block diagram 200 illustrating the one or more control units 104, in accordance with an embodiment of the present disclosure. The one or more control units 104 includes the vehicle control unit 202 and the motor control unit 204, or a combination thereof to monitor the plurality of vehicle driving data variables in real-time or near real-time and to carry out the range estimation. The one or more control unit 104 estimates the remaining range of the vehicle locally on-board the vehicle using the one or more processors 106 as explained in Fig.1.
In some embodiments of the present invention, the one or more control units 104 may independently monitor different driving data variables.
In accordance with an embodiment of the present invention, the vehicle control unit (VCU) 202 is configured to continuously monitor a plurality of driving data variables (historical and current) to estimate the remaining range of the vehicle in real-time or near real-time. Moreover, the plurality of historical and current driving data variables includes, the battery state-of-charge (SoC), the vehicle load, the motor speed, and the current drawn by the motor. Further, the vehicle load estimation algorithm is also performed by the vehicle control unit (VCU) 202 and sends the vehicle load values as an input to the prediction model.
The motor control unit (MCU) 204 is configured to monitor the motor and provide various motor-oriented parameters, such as motor speed and/or motor torque as an input.
Fig. 3 is a flowchart 300 illustrating a method for predicting a remaining driving range for a vehicle in accordance with an embodiment of the present disclosure. The method starts at step 305 and proceeds to step 340. In particular, the method is performed by a range prediction model to predict the remaining driving range for the vehicle. The vehicle is one of a battery electric vehicle (BEV), a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), and a fuel-cell electric vehicle (FCEV). The vehicle is a two-wheeled vehicle or a multi-wheeled vehicle.
At step 305, determining or collecting a plurality of historical data values for a first set of historical driving data variables, and a historical driving range corresponding to the first set of historical driving data variables for a plurality of past trips by one or more processors 106, wherein the plurality of historical data values of the first set of historical driving data variables are collected over multiple time intervals.
At step 310, the plurality of historical data values of each of the historical driving data variables of the first set of historical driving data variables are grouped into a corresponding plurality of bins to form a histogram of each of the historical driving data variables of the first set of historical driving data variables.
At step 315, a historical data value for each of the historical driving data variables of a second set of historical driving data variables is determined.
At step 320, training, by the one or more processors 106, based on the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range, a range prediction model, wherein based on the training, the range prediction model learns a correlation between the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range.
At step 325, determining or collecting a plurality of current data values for a first set of current driving data variables for the ongoing trip by the one or more processors 106. The plurality of current data values of the first set of current driving data variables are collected over multiple time intervals. The multiple time intervals include, but are not limited to, 1 minute, 10 minutes, 1 hour, and 10 hours.
In accordance with an embodiment of the present disclosure, the first and second of historical and current driving data variables includes, but not limited to, a battery state-of-charge (SoC), a vehicle load, a motor speed, and a current drawn by the motor.
In an example, the first set of historical and current driving data variables preferably include the motor speed and the current drawn by the motor. Further, the motor speed is measured by the motor control unit 204 and the current drawn by the motor is measured by the BMS 102. The rest of the two driving data variables i.e. a battery state-of-charge (SoC) and a vehicle load is calculated/measured instantaneously and provided as inputs to the range prediction model.
At step 330, the plurality of current data values of each of the current driving data variables of the first set of current driving data variables are grouped by the one or more processors 106 into a plurality of bins to form a histogram of each of the current driving data variable of the first set of current driving data variables. The lower values of the first set of current driving data variables are grouped into lower order bins and the higher values of the first set of current driving data variables are grouped into higher order bins in the corresponding histograms of the first set of current driving data variables. In addition, the width of each of the bins of the plurality of bins in the histograms are equal.
In an example, the individual bins of the plurality of bins in the corresponding histograms of motor speed and current drawn by the motor include the different range of values for motor speed and current drawn by the motor, respectively, as collected over multiple time intervals. The histograms of motor speed and current drawn by the motor are either stored locally on-board the vehicle, or on the server or the cloud computing platform.
In accordance with an embodiment of the present invention, the lower values of motor speed are grouped into lower order bins in the motor speed histogram and the higher values of motor speed are grouped into higher order bins in the motor speed histogram. In addition, the width of each of the bins of the plurality of bins are equal.
In accordance with an embodiment of the present invention, the lower values of current drawn by the motor are grouped into lower order bins in the histogram of current drawn by the motor, and the higher values of current drawn by the motor are grouped into higher order bins in the histogram of current drawn by the motor.
At step 335, an instantaneous data value for each of the current driving data variables of a second set of current driving data variables for the ongoing trip are determined by the one or more processor 106.
At step 340, the remaining driving range of the vehicle is predicted using the trained range prediction model based on the second set of current driving data variables, and at least one of the histograms of the first set of current driving data variables and the first set of historical driving data variables.
In an example, the input to the trained range prediction model includes the instantaneous value of the remaining battery state-of-charge (SoC) as measured by the BMS 102 in real-time or near real-time, vehicle mass/load as measured/calculated by the VCU 202 in real time or near real-time, and one-minute average of at least one of the histograms of the first set of historical driving data variables and the first set of current driving data variables.
The method further comprises displaying the predicted remaining range of the vehicle to a user or a rider. The predicted remaining range of the vehicle is displayed to the user on a display device 110 including but not limited to one of a CRT display, an LCD display, an LED display, an OLED display, an AMOLED display, and a PMOLED display. In addition, the display device 110 is touch-sensitive or non-touch sensitive.
The trained range prediction system is configured either locally on-board the vehicle, or on a server or a cloud computing platform. Further, the trained range prediction system assigns weights to each of the plurality of input driving data variables. Furthermore, the trained range prediction model assigns weights corresponding to each of the bins of the plurality of bins of the histograms for motor speed and current drawn by the motor, respectively. Also, the trained range prediction model assigns weights to each of the plurality of historical driving data variables corresponding to each of the plurality of driving modes and the driver profiles available in the vehicle. In accordance with an embodiment of the present invention, the trained range prediction model is trained corresponding to at least one of a plurality of driving modes available in the vehicle, at least one of a plurality of driver profiles, and a combination thereof.
In accordance with an embodiment of the present invention, predicting the vehicle range corresponds to a plurality of driver profiles or user profiles available in the vehicle. The driver profile or user profile corresponds to a driving pattern or a driving behaviour of a specific user. For example, if the driver drives rough, the higher bins of the histograms get filled and if the driver drives mellow, the lower bins of the histograms get filled.
In accordance with an embodiment of the present invention, the one-minute average is calculated using one of a moving average, a running average, and an average in general based on the computational ability. The one-minute average of the first set of historical driving data variables is calculated using the plurality of historical data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving. The one-minute average of the first set of current data variables is calculated using the plurality of current data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving.
In addition, the format of the data values for the historical driving data variables and the current driving data variables is one of a real value, an absolute value, and a percentage value. In accordance with an embodiment of the present invention, the method captures the driving data values for the first set and second set of current driving data variables for the ongoing trip based on the selected driver profile and the selected driving mode.
In accordance with an embodiment of the present disclosure, the method further comprises collecting the plurality of current data values over multiple time intervals for the ongoing trip and subsequently, modifying the corresponding existing data values in the histograms of the first set of historical data variables with the current data values of the first set of current driving data variables for the ongoing trip.
Fig. 4 is a flowchart 400 illustrating an exemplary method to predict the remaining driving range of a vehicle with a trained range prediction model in accordance with an embodiment of the present disclosure. The method starts at step 405 and proceeds to step 435.
In an embodiment of the exemplary method, the plurality of historical data values of the motor speed and current drawn by the motor as histograms are pre stored in the vehicle during the vehicle production phase. At step 405, the vehicle is turned ON At step 410, the histograms of motor speed and current drawn by the motor starts developing by collecting, with the help of one or more processors 106 of the one or more control unit 104, the data values of the motor speed and current drawn by the motor for the ongoing trip. The currently captured data values for motor speed and current drawn by the motor subsequently replaces or overwrites the prestored historical data values in the histograms of motor speed and current drawn by the motor. Simultaneously, steps 415 and 420 are implemented in the method in which the battery management system (BMS) 102 measures the battery state of charge value instantaneously, and vehicle control unit (VCU) 202 estimates vehicle load/mass value, respectively. At step 425, the one-minute average of the histograms of motor speed and current drawn by the motor, the instantaneous values of battery SoC and estimated vehicle mass are fed as inputs to a trained prediction model, which further predicts the remaining range of the vehicle as output of step 430. At step 435, the estimated remaining range of the vehicle is displayed to a user on the vehicle display 110 or the display of the user handheld device 112.
In an embodiment of the present invention, collecting the current data values of the motor speed and current drawn by the motor, and subsequently modifying the corresponding existing historical data values of the motor speed and current drawn by the motor with the current data values of the motor speed and current drawn by the motor of the ongoing trip.
In an embodiment of the present invention, the historical data value for SOC and vehicle mass are also used to train the range prediction model.
In an alternate embodiment, the vehicle display device 110 is communicably connected to the user handheld device 112 to display the estimated remaining range of the vehicle.
The system achieves computational efficiency without increasing the overall cost of the product by not requiring additional hardware for implementation. Further, the range prediction system aids in overall efficiency gains in other systems/algorithms using the predicted vehicle range.
, Claims:We Claim:
1. A method for predicting a remaining driving range for a vehicle, the method comprising:
determining, by one or more processors 106, a plurality of historical data values for a first set of historical driving data variables and a historical driving range corresponding to the first set of historical driving data variables for a plurality of past trips, wherein the plurality of historical data values of the first set of historical driving data variables are collected over multiple time intervals;
grouping, by the one or more processors 106, the plurality of historical data values of each of the historical driving data variables of the first set of historical driving data variables into a corresponding plurality of bins to form a histogram of each of the historical driving data variables of the first set of historical driving data variables;
determining, by the one or more processors 106, a historical data value for each of the historical driving data variables of a second set of historical driving data variables;
training, by the one or more processors 106, based on the first set of historical driving data variables and the corresponding historical driving range, and the second set of historical driving data variables, a range prediction model, wherein based on the training, the range prediction model learns a correlation between the first set of historical driving data variables, the second set of historical driving data variables and the corresponding historical driving range;
determining, by the one or more processors 106, a plurality of current data values for a first set of current driving data variables for the ongoing trip, wherein the plurality of current data values of the first set of current driving data variables are collected over multiple time intervals;
grouping, by the one or more processors 106, the plurality of current data values of each of the current driving data variables of the first set of current driving data variables into a corresponding plurality of bins to form a histogram of each of the current driving data variables of the first set of current driving data variables;
determining, by the one or more processors 106, an instantaneous data value for each of the current driving data variables of a second set of current driving data variables for the ongoing trip;
predicting, using the trained range prediction model, the remaining driving range of the vehicle, based on the second set of current driving data variables, and at least one of the histograms of the first set of current driving data variables and the first set of historical driving data variables.

2. The method as claimed in claim 1, wherein the first set and the second set of historical and current driving data variables include at least one of a battery state-of-charge (SoC), a vehicle load, a motor speed, and a current drawn by the motor.

3. The method as claimed in claim 1, wherein during training, the first and the second set of historical driving data variables are used as input variables, and the historical driving range is used as an output variable.

4. The method as claimed in claim 1, wherein the multiple time intervals include, but not limited to, 1 minute, 10 minutes, 1 hour, and 10 hours.

5. The method as claimed in claim 1, wherein the format of the data values for the first and second set of historical driving data variables and the current driving data variables is one of a real value, an absolute value, and a percentage value.

6. The method as claimed in claim 1, wherein the plurality of historical data values of the first set of historical driving data variables are pre stored in the vehicle during the vehicle production phase.

7. The method as claimed in claim 1, further comprises collecting the current data values of the first set of the current driving data variables, and subsequently modifying the corresponding existing historical data values of the first set of historical driving data variables with the current data values of the first set of current driving data variables of the ongoing trip.

8. The method as claimed in claim 1, wherein during prediction, the input to the trained range prediction model includes, instantaneous values of the second set of current driving data variables, and a one-minute average of at least one of the histograms of the first set of current driving data variables and the first set of historical driving data variables.

9. The method as claimed in claim 8, wherein the one-minute average is calculated using one of a moving average, a running average, and an average in general based on the computational ability.

10. The method as claimed in claims 8 and 9, wherein the one-minute average of the first set of historical driving data variables is calculated using the plurality of historical data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving.

11. The method as claimed in claims 8 and 9, wherein the one-minute average of the first set of current driving data variables is calculated using the plurality of current data values collected over multiple preferred time intervals preferably in 1 minute, 10 minutes, 1 hour, and 10 hours of driving.

12. The method as claimed in claim 1, wherein predicting the vehicle range corresponds to a plurality of driving modes available in the vehicle.

13. The method as claimed in claim 12, wherein the plurality of driving modes includes, one of an eco-mode, a ride mode, a rush mode, a sport mode, and a limp mode.

14. The method as claimed in claim 1, wherein predicting the vehicle range corresponds to a plurality of driver profiles or user profiles available in the vehicle.

15. The method as claimed in claim 14, wherein the driver profile or user profile correspond to a driving pattern or a driving behaviour of a specific user.

16. The method as claimed in claim 1, wherein the method captures the driving data values for a first set of current driving data variables for the ongoing trip based on the selected driver profile and the driving mode for the selected driver profile.

17. The method as claimed in claim 1, wherein the trained range prediction model is configured either locally on-board the vehicle, or on a remote computing platform.

18. The method as claimed in claim 1, wherein the range prediction model is trained corresponding to at least one of a plurality of driving modes available in the vehicle, at least one of a plurality of driver profiles available in the vehicle, and a combination thereof.

19. The method as claimed in claim 1, wherein the method is also performed by one or more control units 104 configured with one or more processors 106.

20. The method as claimed in claim 19, wherein the control unit 104 is one of a vehicle control unit (VCU), an electronic control unit (ECU), and a motor control unit (MCU).

21. The method as claimed in claim 1, further comprises displaying the predicted remaining range of the vehicle to a user or a rider.

22. The method as claimed in claim 21, wherein the predicted remaining range of the vehicle is displayed to the user on a display device 110 including, but not limited to, one of a CRT display, an LCD display, an LED display, an OLED display, an AMOLED display, and a PMOLED display.

23. The method as claimed in claim 22, wherein the display device 110 is touch-sensitive or non-touch sensitive.

24. The method as claimed in claim 1, wherein the vehicle is one of a battery electric vehicle (BEV), a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), and a fuel-cell electric vehicle (FCEV).

25. The method as claimed in claim 1, wherein the vehicle is a two-wheeled vehicle or a multi-wheeled vehicle.