Title of Invention: Steel making using laser vibration measurement-Control and condition analysis method of playing process equipment and system using the same
Technical field
[One]
The present invention relates to a system and method for analyzing equipment conditions in a steelmaking-casting process, and more specifically, measuring and diagnosing and analyzing vibrations of equipment of a steelmaking casting process such as an immersion nozzle or a shroud nozzle using a laser, but measuring and analyzing By storing the generated DB and using deep learning, machine learning, and data mining technology to alarm the user of necessary data, it enables real-time monitoring and prediction of the state of the steel-making process equipment.
Background
[2]
The continuous casting process refers to a process in which molten metal, that is, molten metal, is injected into a mold and cooled while continuously manufacturing metal casts such as blooms, billets, and slabs. The shroud nozzle is used to prevent reoxidation due to atmospheric contact when casting molten steel between tundish in a ladle. Whenever the ladle is replaced, the shroud nozzle is disengaged by the manipulator arm and rejoined prior to casting.
[3]
The molten metal moves from the ladle to the tundish through the shroud nozzle and is injected into the mold from the tundish through the immersion nozzle. The flow of molten metal flowing into the mold and the initial solidification process are one of the factors that determine the quality of the metal cast that has been continuously cast.
[4]
However, if the combination is abnormally performed during the separation-recombination process of the shroud nozzle, the local drop impact is concentrated on the inner wall of the inclined nozzle, causing cracks and breakage, or the mixing of outside air into the nozzle, resulting in a problem of lowering the quality of cast steel. Occurs.
[5]
Such a twisting of the nozzle verticality can act as a risk factor for safety accidents in the continuous casting process, but at present, no appropriate measures have been taken. In addition, there is a problem that the slag floating inside the ladle due to the difference in specific gravity causes a vortex by the turning force at the end of casting and flows into the tundish. Since this adversely affects the strong cleanliness, a technique is required to block the sliding gate valve by measuring the point of time when the slag flows into the shroud nozzle.
[6]
There are two technologies for detecting the inflow of slag so far, the method of measuring the timing of inflow of slag through the change of the electromagnetic field by using the property that the electrical conductivity of the river passing through the shroud nozzle changes as the slag is introduced, and the separation and combination of the shroud nozzle. There is a method of detecting the vortex vibration occurring at the time of inflow of slag by attaching an acceleration vibration sensor to the manipulator arm.
[7]
However, these existing technologies have many problems. The electromagnetic field change measurement method is susceptible to heat because the sensor is inserted inside the nozzle, causing frequent failures, and the attached vibration measurement method using an acceleration sensor measures the attenuated vibration through a manipulator, resulting in poor measurement accuracy.
[8]
In addition, in some cases, the shroud nozzle disposed between the ladle and the tundish may have defects such as cracks or fractures of the nozzle due to poor verticality. Such defects can lead to a decrease in the quality of cast steel due to the inflow of outside air, and of molten steel. Problems can arise continuously, which in turn cause further damage to the nozzle due to partial impact concentration.
[9]
In order to solve these problems, the applicant developed a non-contact type vibration measurement system using a laser vibration measurement technology according to the present application. This analysis system analyzes and checks the state of steelmaking-casting process facilities, analyzes the state of the immersion nozzle, the verticality of the shroud nozzle or immersion nozzle is distorted, the time of inflow of the slag of the shroud nozzle, and the occurrence of cracks in the shroud nozzle in real time. It provides information to the operator to monitor the status of the performance by wire or wirelessly.
[10]
In addition, various defects may occur in the immersion nozzle other than the shroud nozzle.In some cases, the immersion nozzle through which the molten steel passes through the slag inflow into the ladle, the deoxidizable inclusions mixed in the molten steel, or the oxide formed by the reaction between the refractory material of the shroud nozzle and the molten steel. Nozzle clogging may occur as the inner diameter decreases. When such a clogging of the immersion nozzle occurs, the flow of molten steel is not constant, so that the height of the molten steel becomes unstable, and the quality of the continuous casting product is adversely affected by the change of the hot water surface. In addition, drift occurs due to clogging of the immersion nozzle, and mold flux or the like is mixed into the molten steel, thereby deteriorating the cleanliness of the molten steel.
[11]
Moreover, when a part of the clogging material (inclusions or oxides) adhering to the inner wall of the immersion nozzle falls off, the fallen part is mixed in the molten steel and deteriorates the cleanliness of the molten steel. In case of severe clogging of the immersion nozzle, the casting process may be stopped.
[12]
As a method of detecting defects in the shroud nozzle or immersion nozzle in the steel-making-casting process facility, the detection method through the opening rate of the sliding gate during the continuous casting process, and the operator contacting the steel rod with the nozzle wall surface with the touch of the hand. There is a detection method through the obtained vibration.
[13]
However, in the method of detecting the degree of clogging of the nozzle according to the opening rate of the sliding gate, it is difficult to detect the clogging of the immersion nozzle until the inside of the nozzle is clogged by about 50% or more. In addition, the method in which the operator touches an iron rod to detect vibration with the touch of his hand has limitations in that there is a difference between operators, lack of accuracy, and difficulty in real-time online measurement.
[14]
In order to solve this problem, Publication No. 10-2009-0071221 "Method and device for predicting clogging of immersion nozzle" is attempting to analyze the state of immersion nozzle by measuring the frequency of immersion nozzle. Since it provides only a simple prediction of the state, the utilization of the DB is very low, and there is a problem that the possibility of application in the field is low because there is no function of providing real-time information or an alarm to the user.
[15]
In addition, the steelmaking-playing process is a very dangerous and difficult to access environment, so it is very important for the user or operator to check and inspect the facilities or systems of the above process, but it is having difficulty in proceeding. Even if the nozzle itself is defective or maintenance is required, it is difficult to check the condition and access for maintenance due to the working environment of high temperature, high temperature, and high pressure.
Detailed description of the invention
Technical challenge
[16]
The present invention has been conceived in view of the above-described problems, and an object of the present invention is to measure and analyze the frequency of steelmaking-playing process facilities using a laser, but use deep learning and data mining technologies to utilize the collected and analyzed DB. It provides an analysis system and method capable of preparing big data by utilizing it, analyzing and predicting through learning of artificial intelligence (AI), and real-time alarm and display in the field.
Means of solving the task
[17]
The method of analyzing the state of the steelmaking-performing process facility according to the present invention includes the steps of measuring the vibration data of the steelmaking-performing process facility using a laser vibration meter, transmitting the measured vibration data, the received measured vibration data and the preset Comparing and analyzing data, storing and learning data, and displaying the compared and analyzed data, wherein the measured data is analyzed for correlation by case or preset defects according to each situation condition and use state. It determines whether there is an abnormality in the facility or process through a diagnostic algorithm, sends an alarm or control signal if necessary, converts the data into a DB (database), but stores it as big data by using data mining, machine learning, or deep learning technology. It is characterized in that the user can monitor and diagnose the state of the steelmaking-performing process facility in real time by displaying the utilized, compared and analyzed data to the user.
[18]
In addition, the storing and learning of the data includes converting real-time vibration information into DB, converting abnormal conditions occurring in the facility into DB, and in the case of nozzles for continuous casting, external air inflow, nozzle fastening, clogging of inner diameter, slag. The step of converting information on the state of mixing, cracking or dropping out of large inclusions into a DB, converting information about correlation analysis between vibration information and state into a DB, converting information about correlation analysis between operating conditions and conditions into a database, and storing information on stability standards It characterized in that it comprises a step of making.
[19]
In addition, the displaying of the compared and analyzed data may further include displaying the data to the user, and providing separate alarm and replacement information to the user when a preset condition is not satisfied.
[20]
In addition, the DB is used as big data, and the predictability of information and the reliability of monitoring data are increased through repetitive learning by applying artificial intelligence technology.
[21]
The system for analyzing the state of the steel-making process equipment according to the present invention comprises: a vibration measuring unit for continuously measuring the surface of the steel making-casting process equipment; A data transmission unit for transmitting the measured data; A data analysis unit that analyzes the measured data; A data storage unit for storing the measured and analyzed data or database; And a data output unit outputting the measured and analyzed data. Including, wherein the data analysis unit determines whether there is an abnormality in the facility or process through a correlation analysis for each case or a preset defect diagnosis algorithm according to each situation condition and use condition, and sends an alarm or control signal if necessary , The data is converted into a DB (database), but it is stored and utilized as big data by using data mining, machine learning, or deep learning technology, and the compared and analyzed data is displayed to the user, so that the user can It is characterized in that it can monitor and diagnose the status in real time.
[22]
In addition, the data analysis unit converts real-time vibration information of the facility into a DB, converts abnormal conditions occurring in the facility into a DB, and in the case of a nozzle for continuous casting, external air inflow, nozzle fastening, clogging of inner diameter, slag mixing, cracking, or It is characterized by converting large inclusions dropping status information into DB, converting vibration information and condition correlation analysis information into DB, working condition and condition correlation analysis information into DB, and use stability reference information into DB.
[23]
In addition, the data output unit displays the data to the user, and provides separate alarm and replacement information to the user when a preset condition is not satisfied.
[24]
In addition, the DB is used as big data, and the predictability of information is increased through repetitive learning by applying artificial intelligence technology.
[25]
In addition, the vibration measuring unit continuously measures the vibration data of the facility by using a non-contact means, and a laser vibration sensor is used, and further comprising a vibration isolation and heat dissipation means.
[26]
In addition, storing and learning the data; The step of converting the real-time vibration information into a DB; Converting an abnormal state occurring in the facility into a DB; In the case of a nozzle for continuous casting, converting information on the state of external air inflow, nozzle fastening, inner diameter clogging, slag mixing, crack occurrence, or large inclusions dropping out into a DB; Converting vibration information and state correlation analysis information into DB; Converting information on correlation analysis between operating conditions and conditions into a DB; And converting the use stability reference information into a DB. It characterized in that it comprises a.
[27]
In addition, it is characterized in that the predicted information is transmitted by wire or wirelessly.
[28]
In addition, it transmits control signals or information in connection with the flow control device and steel-making process equipment, and in the case of a nozzle for continuous casting, when an abnormality is detected in the connection of the nozzle, an automatic nozzle position adjustment signal is sent, and when large inclusions fall off, It is characterized by sending information about the time point as a warning signal, and sending a control signal or a replacement signal to the flow control device when a slag sink vortex and slag inflow is detected, or when an inner diameter blockage exceeds a preset value.
[29]
In addition, the data analysis unit converts real-time vibration information of the facility into a DB and converts the abnormal state occurring in the facility into a DB. In the case of nozzles for continuous casting, external air inflow, nozzle fastening, inner diameter clogging, slag mixing, cracking or large It is characterized in that the information on the state of dropping inclusions is converted into a DB, information on the correlation analysis between operating conditions and conditions is converted into a DB, and information on the use stability standard is converted into a DB.
[30]
In addition, it is characterized in that the monitoring information calculated through the analysis unit is transmitted using a wired, wireless, or IoT environment.
[31]
In addition, it characterized in that it further comprises a function of fixing the measurement position by digital image processing of the camera.
[32]
In addition, it transmits control signals or information in connection with the flow control device and steel-making process equipment, and in the case of a nozzle for continuous casting, when an abnormality is detected in the connection of the nozzle, an automatic nozzle position adjustment signal is sent, and when large inclusions fall off, The time information is sent as a warning signal, and when the inflow of slag sink vortex and slag is detected, or the blockage of the inner diameter exceeds a specific value, a control signal or a replacement signal is sent to the flow control device.
[33]
In addition, the vibration measurement unit continuously measures the vibration data of the steelmaking-performing process facility by using a non-contact means, and a laser vibration meter is used. Diagnosis of inflow, nozzle fastening, clogging of inner diameter, slag mixing, cracking, or dropping of large inclusions, and efficient steel making by transmitting control signals to peripheral devices based on the diagnosis contents-Improvement of cast quality and molten steel through the control of the playing process It is characterized by increasing the error rate.
[34]
In addition, the steelmaking-playing process facility is characterized in that it is a shroud nozzle or an immersion nozzle.
Effects of the Invention
[35]
The system and method for state analysis of steelmaking-performing process equipment according to the present invention measure and analyze the frequency of an object using a laser vibration meter, but utilize a DB analyzed using deep learning and machine learning technologies, and, for example, generated from a nozzle. It predicts nozzle defects, warping, cracking, mixing of outside air, and slag inflow, and enables real-time alarms and displays in the field.
Brief description of the drawing
[36]
1 is a schematic diagram showing a continuous casting apparatus according to the present invention,
[37]
Figure 2 is an explanatory diagram for diagnosing the state of the shroud nozzle, one of the steelmaking-casting process equipment according to the present invention,
[38]
3 is a cross-sectional explanatory view of a normal state of the shroud nozzle according to the present invention
[39]
4 is a cross-sectional explanatory view of a state in which a defect occurs in the shroud nozzle according to the present invention;
[40]
5 is an explanatory diagram of analysis of a vibration factor of the shroud nozzle according to the present invention;
[41]
6 is a graph of vibration measurement of the shroud nozzle in a steady state according to FIG. 3;
[42]
7 is a graph of vibration measurement of a shroud nozzle in a defect occurrence state according to FIG. 4;
[43]
8 is a block diagram of a system for analyzing a condition of a shroud nozzle according to the present invention,
[44]
9 is a flow chart of a first embodiment of a method for analyzing a state of a shroud nozzle according to the present invention;
[45]
10 is a flow chart of a second embodiment of a method for analyzing a state of a shroud nozzle according to the present invention;
[46]
11, 12 and 13 are partial detailed views of a method for analyzing a state of a nozzle according to the present invention,
[47]
14 is an explanatory diagram of a system and method for analyzing a state of a shroud nozzle according to the present invention,
[48]
15 and 16 are field photos and display related photos for testing the system and method for analyzing the condition of the shroud nozzle according to the present invention,
[49]
17 and 18 are graphs of vibration measurement results in a normal state and a defective state according to the test,
[50]
19 and 20 are internal photographs of a normal state and a defective state of the shroud nozzle according to the present invention,
[51]
21 and 22 are graphs of vibration measurement results in a normal state and a defective state,
[52]
23 and 24 are reference diagrams for explaining the peripheral relationship of the shroud nozzle according to the present invention;
[53]
25 is an explanatory diagram of a shroud nozzle manipulator applied to the present invention;
[54]
26 is an explanatory diagram illustrating a state of diagnosing the state of the immersion nozzle, which is another of the steelmaking-casting process equipment according to the present invention;
[55]
27 is an explanatory diagram of the configuration of a system for measuring the state of the immersion nozzle according to the present invention.
Best mode for carrying out the invention
[56]
Hereinafter, with reference to the accompanying drawings, with respect to the control, condition analysis system and method of the steel making-performing process equipment according to an embodiment of the present invention, in detail so that those skilled in the art can easily implement Let me explain. The present invention is not limited to the embodiments described herein, and may be implemented in various different forms. In order to clearly describe the present invention, parts irrelevant to the description have been omitted, and the same reference numerals are attached to the same or similar components throughout the specification.
[57]
[58]
Throughout the specification of the present invention, the molten metal refers to a liquid state in which the metal is melted during the casting operation. Since it refers to a state in which the metal is dissolved, it is solidified after cooling to change into a metal cast. In this case, the metal cast may be a slab, bloom, or billet. Specifically, when casting steel, the molten metal may refer to molten steel.
[59]
[60]
In addition, throughout this specification, data and databases that refer to information values ​​may be used with the same meaning depending on circumstances, and if a new meaning occurs through accumulation and storage of data, the existing data may be referred to as a database. will be.
[61]
[62]
Steel making to measure vibration and analyze the state by using a laser according to the present invention-Equipment and systems of the playing process are not limited to the devices or facilities described later, and include devices, facilities, and systems in all processes, preferably Refers to shroud nozzles and immersion nozzles.
[63]
[64]
1 is a schematic diagram showing a continuous casting apparatus for the practice of the present invention. As shown, the molten metal 20 in the ladle 70 moves to the tundish 10 through the shroud nozzle 200, and the tundish 10 and the tundish received molten metal from the ladle 70 The immersion nozzle 100 is connected to the bottom surface of (10), and the immersion nozzle 100 is inserted into the mold 30 defining the shape of the cast steel.
[65]
[66]
2 is a partial view for explaining the state analysis system and method of the shroud nozzle 200, which is one of the facilities of the steelmaking-playing process according to the present invention. As shown, the shroud nozzle 200 may be connected to the ladle 70 by further providing a collector nozzle 220 and an SN plate 210 thereon, and being spaced apart from the shroud nozzle 200 Vibration measuring units 300 and 310 for observing and detecting vibration of the nozzle 200 are provided.
[67]
[68]
The molten metal 20 moves from the ladle 70 to the tundish 10 through the shroud nozzle 200, and the verticality of the shroud nozzle 200 is unbalanced or unbalanced, or cracks or fractures occur in the nozzle. If this occurs, external air may be introduced into the gap, thereby reducing reoxidation of molten steel, causing an abnormality in the cleanliness of the steel, and the problem of inflow of slag of the ladle 70 may be repeated.
[69]
[70]
In addition, the molten metal 20 is injected into the mold 30 from the tundish 10 through the immersion nozzle 100 to form a solidification layer and undergo an initial solidification process. The solidified layer exiting the mold 30 is cooled by a cooling medium sprayed through a cooling unit, such as a spray nozzle, to form a metal cast in the shape of a slab, bloom, or billet.
[71]
[72]
Subsequently, the metal cast is moved to the next step by a guide roller. In addition, it is possible to cut the cast piece moving at the cutting point according to the desired length and size. The flow of the molten metal 20 flowing into the mold 30 and the initial solidification process are important factors that influence the properties and quality of the metal cast that has been continuously cast. Accordingly, the present invention detects, analyzes, and measures defects in the shroud nozzle 200 to prevent or respond to them, thereby improving the quality of the molten metal and the cast iron.
[73]
[74]
As such, the quality of the cast iron may be reduced due to defects or abnormalities of the shroud nozzle 200, so it is necessary to promptly and continuously detect and analyze the state of the shroud nozzle 200 to respond and alarm the user.
[75]
[76]
3 is an enlarged cross-sectional view including the shroud nozzle 200 according to the present invention, showing a normal state, and as shown, the non-contact vibration measuring unit 300 vibrates the shroud nozzle 200 at a distance. Measure the data. In addition, it can be seen that the SN plate 210, the collector nozzle 220, and the shroud nozzle 200 are normally connected so that external air does not flow.
[77]
[78]
6 is a graph showing data as a result of measuring the vibration of the shroud nozzle 200 in a normal state as described above. When there is no defect, a particularly high-amplitude vibration does not occur, and an even state can be seen.
[79]
[80]
4 is an enlarged cross-sectional view including the shroud nozzle 200 according to the present invention, showing an abnormal state, that is, a state in which a defect or an abnormality has occurred, and as shown, the non-contact vibration measuring unit 300 is The vibration data of the shroud nozzle 200 is measured at. In addition, it can be seen that the SN plate 210, the collector nozzle 220, and the shroud nozzle 200 are connected abnormally so that the inflow of external air occurs.
[81]
[82]
As described above, if the verticality of the shroud nozzle 200 is poor, external air is introduced and the quality of the cast iron is deteriorated. In some cases, cracks may occur in the shroud nozzle 200. When molten steel 20 is introduced in such an abnormal state, abnormal vibration data is generated in the shroud nozzle 200.
[83]
[84]
7 is a graph showing data as a result of measuring the vibration of the shroud nozzle 200 in an abnormal state as described above. When there is a defect, vibration of a high amplitude occurs, and as shown in the figure, at a frequency of 6 to 8 Hz. It can be seen that very high amplitude vibrations can occur. In this way, it is possible to check or predict the presence or absence of a nozzle abnormality and status information in real time using vibration data of the shroud nozzle 200.
[85]
[86]
5 is a view for explaining the analysis of the vibration factor or factor of the shroud nozzle 200, the flow rate change of the molten metal in the ladle 70, the precipitation or rotational flow information of the slag, the opening rate of the SN plate 210 It can comprehensively detect, analyze, and judge various factors such as drift and inflow of external air at the connection.
[87]
[88]
Since the vibration data that changes due to various factors can be measured, an algorithm through weight or variable application for each factor should be applied to the vibration data obtained when analyzing the data, and the obtained data is processed and used as a reference value afterwards. It might be possible.
[89]
[90]
The algorithm is provided as a data-based defect diagnosis algorithm by analyzing the measured vibration data, and at this time, correlation analysis and nozzle defect determination criteria are established by linking the vibration change data with the defect status check data of the shroud nozzle after use and operating conditions. Can be prepared.
[91]
[92]
As shown in FIG. 2, the system and method for analyzing the condition of the shroud nozzle according to the present invention measures the vibration of the shroud nozzle 200, compares and analyzes the measured data, and stores the data as a database. It is characterized in that real-time monitoring is possible by alarming and outputting and predicting the state of the shroud nozzle 200 when necessary.
[93]
[94]
As shown in FIG. 9, the method for analyzing the state of the shroud nozzle according to the present invention includes the steps of measuring vibration data of the shroud nozzle surface (S10), storing the measured vibration data (S20), and the received measured vibration data. Comparing and analyzing preset data, preprocessing and converting (S30), determining whether there is a defect (S40), displaying data (S50), and alarming (S60).
[95]
[96]
10 is another embodiment of a method for analyzing the state of a shroud nozzle according to the present invention, a vibration measurement step (S100) of a shroud nozzle surface, a step of transmitting vibration measurement data (S200), a comparison and analysis step of data (S300), It includes storing and learning data (S400), outputting necessary data (S500), and monitoring, learning, and prediction (S600).
[97]
[98]
The storing and learning of the data (S400) comprises: converting real-time vibration information into DB (S410), converting defect state information into DB (S420), converting vibration information and state correlation analysis information into DB ( S430), a step of converting the correlation analysis information between the operating condition and the state into a DB (S440), and a step of converting the use stability reference information into a DB (S450).
[99]
[100]
That is, it is possible to analyze and data mining for each case of the use condition of the shroud nozzle 200, convert the vibration change of the real-time (continuous) shroud nozzle into a DB, check the shroud defect status and convert it into a database after use, and use each shroud nozzle. Analysis of the correlation between the condition after use and the vibration generated at the time of use, analysis of the correlation between the condition after use of the shroud nozzle and the operating conditions (melt steel temperature, casting speed, molten steel composition, etc.), the measured value of vibration of the shroud nozzle and the condition after use, Through the comparative analysis of the correlation of the operating conditions, the standard range setting for the stability of use for the occurrence of defects in the shroud nozzle can be made into a DB (predicted calculation of the occurrence of defects for each case).
[101]
[102]
Displaying the compared and analyzed data (S500) may further include displaying the data to a user, but providing separate alarm and replacement information to the user when a preset condition is not satisfied (S530). .
[103]
[104]
In other words, during operation, the vibration measurement data of the shroud nozzle is output and displayed so that it can be displayed visually.When data outside the preset use stability range is measured, information along with an alarm notifying the risk of using the shroud nozzle and the need to replace it. It is possible to provide a system that detects the degree of defects in the shroud nozzle and changes in vibration measurement data and grasps the degree in real time in advance. Both visual and audio means may be used as the alarm means.
[105]
[106]
In addition, the vibration measuring unit continuously measures the vibration data of the shroud nozzle using a non-contact means, but uses a laser vibration meter, diagnoses nozzle fastening failure, ladle slag inflow, and crack occurrence through data mining, and diagnosis details Based on this, control signals can be transmitted to peripheral devices to increase the error rate of molten steel through efficient gate blocking.
[107]
[108]
In addition, the measured data is converted into a DB (database) through case-by-case correlation analysis according to context conditions and usage conditions, and converted into big data by using data mining, deep learning, or machine learning technology Can be stored and utilized.
[109]
[110]
The data mining to be applied to the state analysis method according to the present invention is a process of finding knowledge of interest by any method (sequential pattern, similarity, etc.) in a database, or a process of finding useful information in a large amount of data. It refers to the technology that allows you to find information that you did not expect as well as information you did.
[111]
[112]
By grasping the relevance of information through data mining, you can maximize profits by creating valuable information and applying it to decision-making. Discover hidden knowledge, unexpected trends or new rules based on all available source data, including, for example, daily transaction data, customer data, product data, or other external data other than customer response data from various marketing activities. In other words, it is intended to be used as information for actual business decisions.
[113]
[114]
The most representative application field of data mining is database marketing. Accordingly, the surface vibration data of the shroud nozzle according to the present invention can be processed and classified and used as a database, and the data mining is applied thereto.
[115]
[116]
The deep learning applied to the method of analyzing the state of the shroud nozzle according to the present invention is a machine learning technology built on the basis of an artificial neural network (ANN) in order to enable a computer to learn by itself like a human using various data. Say.
[117]
[118]
In deep learning, the human brain learns a machine to discern objects by imitating the information processing method that distinguishes objects after discovering patterns in a number of data. When deep learning technology is applied in this way, computers can recognize, reason, and judge by themselves even if a person does not set all judgment criteria.
[119]
[120]
Therefore, deep learning technology is widely used for voice, image recognition, and photo analysis. Therefore, since the information on the state of the shroud nozzle is arranged and analyzed, and the computer can recognize, infer, and judge by itself through this, convenience in use is provided by allowing the user to predict the state of the shroud nozzle.
[121]
[122]
[123]
In this way, if the condition of the shroud nozzle and whether or not defects can be predicted, it is possible to procure, manage, and prepare for stock when necessary, thereby increasing productivity and obtaining excellent effects in terms of time and economy.
[124]
[125]
In addition, by displaying the compared and analyzed data to the user in real time, the user can monitor the state of the shroud nozzle in real time. At this time, the display method can be determined by a direct alarm to the user or through a comprehensive control room, and when an alarm is directly sent to the user, the alarm can also be performed through the user's portable terminal, PDA, smartphone, or tablet PC.
[126]
In the present invention, as described above, an alarm for the state information of the shroud nozzle is possible to the user in real time, so that the user can intuitively grasp the current state, and through this, it is possible to prepare for prediction and response.
[127]
[128]
The above data or DB is used as big data, but the IoT technology can be applied to increase the predictability of repetitive learning and information through the application of artificial intelligence technology, and each of the components prepared in the field to transmit and receive data to each other. .
[129]
[130]
That is, the manager at the site, the manager's portable terminal (not shown), and each component of the workplace provide and receive information to each other, enabling efficient, quick and accurate analysis and prediction compared to the existing shroud nozzle condition analysis method. will be.
[131]
[132]
Big data used in accordance with the present invention refers to data generated in a digital environment and has a large scale, a short generation period, and a form of large-scale data including not only numerical data but also text and image data.
[133]
[134]
8 is a block diagram of a configuration of a system for analyzing a condition of a shroud nozzle according to the present invention. The system for analyzing a condition of a shroud nozzle according to the present invention includes a vibration measuring unit 300 for continuously measuring the surface of the shroud nozzle, and the measured data. A data transmission unit 400 to transmit, a data analysis unit 500 to analyze the measured data, a data storage unit 700 to store the measured and analyzed data or a database, and output the measured and analyzed data It includes a data output unit 600.
[135]
[136]
The data analysis unit 500 converts the measured data into a DB (database) through case-by-case correlation analysis according to each situation condition and use state, and stores it as big data by using data mining, machine learning or deep learning The utilized, compared and analyzed data is displayed to the user in real time, so that the user can monitor the status of the shroud nozzle in real time.
[137]
[138]
In addition, the data analysis unit 500 converts real-time vibration information into a DB, converts defect status and defect occurrence time, and crack status information into a DB, converts vibration information and condition correlation analysis information into a DB, and analyzes the correlation between operating conditions and conditions. The information can be converted into a database and the information on the stability standard can be converted into a database.
[139]
[140]
In addition, the storing and learning of the data includes the steps of converting real-time vibration information into DB, nozzle coupling distortion, external air mixing, crack generation, and slag inflow state information into DB, and correlation analysis information between vibration information and state. It may include converting DB, converting information about correlation analysis between operating conditions and conditions into DB, and converting information about stability criteria into DB. In addition, the predicted information can be delivered by wire or wirelessly.
[141]
[142]
In addition, it transmits the predicted diagnosis signal or information in connection with the ladle plate mechanism and manipulator-arm, sends a signal to automatically adjust the nozzle position when a nozzle tightening error is detected, and sends a mechanical open/close signal when a slag inflow is detected from the ladle. In this way, it may further include a step or system for efficiently establishing a casting blocking point.
[143]
[144]
In addition, the data output unit 600 displays data to the user, but may provide separate alarm and replacement information to the user when a preset condition is not satisfied.
[145]
[146]
In addition, although the DB is used as big data, it is possible to apply artificial intelligence technology to increase the predictability of repetitive learning and information through it, and IoT technology that enables each component prepared in the field to transmit and receive data to each other can be applied. In some cases, such as security facilities, information can be transmitted and received over the wire.
[147]
[148]
15 shows a test process for confirming the state analysis method and system of the shroud nozzle according to the present invention, and the vibration measurement unit 300 measured at a distance of 10 meters or more from the test part using a laser. , In some cases, it may be measured using infrared rays. This test was conducted in an actual field, and the vibration measuring unit 300 was measured at a distance of 7m from the shroud nozzle 200.
[149]
[150]
16 is a screen for monitoring the point of the shroud nozzle to be measured by the vibration measuring unit 300 as described above. In addition to the laser vibration sensor, various devices and means, such as an infrared photographing device and a temperature measuring means using a long distance, can be simultaneously applied. .
[151]
[152]
In addition, in addition to the non-contact sensor, an image photographing device (not shown) may be further provided in order to increase the accuracy of measurement, and information such as changes in the external state, position, and color of the shroud nozzle may be provided through the image photographing device. It can be detected and compared at the same time.
[153]
[154]
The vibration measuring unit 300 may be provided with a separate anti-vibration and heat dissipation means, and the vibration-isolating means may use a housing, a case, a cover, or a damper to prevent vibration due to external factors of the vibration measuring unit 300 The heat dissipation means may be provided with a fan or the like to prepare for a high-temperature environment acting near the shroud nozzle 200.
[155]
In addition, the vibration measuring unit 300 may include a vibration isolation, heat dissipation means, and a measurement position tracking function by image processing, and an auto tracking function may be included to measure the same position. In other words, because the shroud nozzle is moved by the opening and closing rate of the sliding gate valve during operation, the measurement position is moved. will be.
[156]
[157]
17 is a graph representing the vibration measurement result of the shroud nozzle in a steady state according to the above test result. No vibration in amplitude is detected.
[158]
[159]
[160]
However, as shown in FIG. 18, the vibration measurement graph in an abnormal state is that external air is introduced, and accordingly, vibration of a very high amplitude is generated in a very specific area. As shown, vibrations having an amplitude of 2 to 4 times that of the normal state occurred at 6 to 8 Hz.
[161]
[162]
In addition, in a normal state, that is, when no defect occurs, the amplitude was within 0.001m in the 0 to 38 Hz section, and when an abnormal state, that is, when a defect occurs, high amplitude vibration was observed in the 0 to 38 Hz section.
[163]
[164]
19 is a photograph showing the inside of the nozzle in a normal state, and FIG. 20 is a photograph showing the inside of the nozzle in an abnormal state of poor verticality. As shown, in the normal state, the accumulation amount on the inner surface of the nozzle is constant, but in the abnormal state, it can be seen that the accumulation amount deviation is about 10 mm.
[165]
[166]
FIG. 21 is a graph related to the vibration measurement result in the normal state according to FIG. 19, and FIG. 22 is a graph related to the vibration measurement result in the abnormal state according to FIG. 20. It can be seen that vibration occurs.
[167]
[168]
23, 24 and 25 are explanatory diagrams for a configuration and an apparatus for measuring the vibration of the shroud nozzle according to the present invention, and FIGS. 26 and 27 are a steelmaking-making process facility according to the present invention, which is another immersion It is a diagram for explaining the state analysis of the nozzle.
[169]
[170]
That is, the steelmaking-performing process equipment control, condition analysis method, and condition analysis system according to the present invention can be used throughout the process equipment, and are preferably applied to the above-described shroud nozzle and immersion nozzle to determine the state by laser vibration measurement. Analyze, predict.
[171]
[172]
For example, shroud nozzles and immersion nozzles have similar purposes to guide the flow of molten steel and prevent oxidation, but the types of defects that may occur may be different depending on the order of the process and the environment. Results can be different.
[173]
[174]
Since the present invention aims to measure and analyze the state of all configurations of the steel making-performing process equipment and utilize the collected data as above, the present invention can be evenly utilized in the entire process, and in some cases, it can be used in various ways other than the steel making-performing process. It could also be applied to fields and processes.
[175]
[176]
The present invention described above is not limited by the above-described embodiments and the accompanying drawings, and simple substitutions, modifications, and changes within the technical spirit of the present invention are apparent to those of ordinary skill in the art.
Industrial applicability
[177]
The system and method for analyzing the state of the steelmaking-performing process facility according to the present invention measure the frequency of the process facility, shroud nozzle or immersion nozzle, compare and analyze the measured data, but prepare big data using deep learning and data mining technologies. And it can be used in the method and system for state analysis of steel-making process facilities that can analyze and predict the situation through learning of artificial intelligence (AI).
[178]
Claims
[Claim 1]
Steel making-measuring the vibration data of the playing process equipment using a laser vibration meter; Transmitting the measured vibration data; Comparing and analyzing the received measured vibration data and preset data; Storing and learning data; And displaying the compared and analyzed data. Including, wherein the measured data is determined whether there is an abnormality in the facility or process through a correlation analysis for each case or a preset defect diagnosis algorithm according to each situation condition and use condition, and sends an alarm or control signal if necessary, and the data DB (database), but by using data mining, machine learning or deep learning technology to store and utilize as big data, and display the compared and analyzed data to the user, the user can view the status of the steelmaking-performance process equipment in real time. Steelmaking-control and condition analysis method of a performance process facility, characterized in that monitoring and diagnosis.
[Claim 2]
The method of claim 1, further comprising: storing and learning the data; The step of converting the real-time vibration information into a DB; Converting an abnormal state occurring in the facility into a DB; In the case of a nozzle for continuous casting, converting information on the state of external air inflow, nozzle fastening, inner diameter clogging, slag mixing, crack occurrence, or large inclusions dropping out into a DB; Converting vibration information and state correlation analysis information into DB; Converting information on correlation analysis between operating conditions and conditions into a DB; And converting the use stability reference information into a DB. Steel making-control and state analysis method of the performance process equipment, characterized in that it comprises a.
[Claim 3]
The method of claim 1, further comprising: displaying the compared and analyzed data; Displaying the data to the user, but providing separate alarm and replacement information to the user when the preset condition is not satisfied; Steel making-control and state analysis method of the performance process equipment, characterized in that it further comprises.
[Claim 4]
The steelmaking-making process facility control and state analysis according to claim 1, wherein the DB is used as big data, and the predictability of information and the reliability of monitoring data are increased through repetitive learning by applying artificial intelligence technology. Way.
[Claim 5]
Steel making-vibration measuring unit for continuously measuring the surface of the playing process equipment; A data transmission unit for transmitting the measured data; A data analysis unit that analyzes the measured data; A data storage unit for storing the measured and analyzed data or database; And a data output unit outputting the measured and analyzed data. Including, wherein the data analysis unit determines whether there is an abnormality in the facility or process through a correlation analysis for each case or a preset defect diagnosis algorithm according to each situation condition and use condition, and sends an alarm or control signal if necessary, , The data is converted into a DB (database), but it is stored and utilized as big data by using data mining, machine learning, or deep learning technology, and the compared and analyzed data is displayed to the user, thereby allowing the user to Steel making-control and condition analysis system of a performance process facility, characterized in that it can monitor and diagnose the condition in real time.
[Claim 6]
The method of claim 5, wherein the data analysis unit converts real-time vibration information of the facility into a DB, converts abnormal conditions occurring in the facility into a DB, and in the case of a nozzle for continuous casting, external air inflow, nozzle fastening, clogging of inner diameter, and slag mixing. , Crack occurrence or large inclusions dropping status information is converted into DB, the correlation analysis information between vibration information and condition is made into DB, the correlation analysis information between operating conditions and condition is made into DB, and the use stability standard information is made into DB Steel making-Control and condition analysis system of performance process equipment.
[Claim 7]
The steelmaking-performing process facility control and status of claim 5, wherein the data output unit displays data to the user, but provides separate alarm and replacement information to the user when a preset condition is not satisfied. Analysis system.
[Claim 8]
The system of claim 5, wherein the DB is used as big data, and artificial intelligence technology is applied to increase the predictability of information through repetitive learning.
[Claim 9]
According to claim 5, wherein the vibration measuring unit continuously measures the vibration data of the facility using a non-contact means, using a laser vibration sensor, and further comprising a vibration-proof and heat dissipation means. Process equipment control and condition analysis system.
[Claim 10]
The method of claim 1, further comprising: storing and learning the data; The step of converting the real-time vibration information into a DB; Converting an abnormal state occurring in the facility into a DB; In the case of a nozzle for continuous casting, converting information on the state of external air inflow, nozzle fastening, inner diameter clogging, slag mixing, crack occurrence, or large inclusions dropping out into a DB; Converting vibration information and state correlation analysis information into DB; Converting information on correlation analysis between operating conditions and conditions into a DB; And converting the use stability reference information into a DB. Steel making-control and state analysis method of the performance process equipment, characterized in that it comprises a.
[Claim 11]
The method of claim 4, wherein the predicted information is transmitted by wire or wirelessly.
[Claim 12]
5. Steelmaking characterized by sending information about the time point as a warning signal when inclusion dropout is detected, and sending a control signal or replacement signal to the flow control device when the inflow of slag sink vortex and slag is detected, or when the inner diameter blockage exceeds a preset value. -Control and condition analysis method of the playing process equipment.
[Claim 13]
The method of claim 5, wherein the data analysis unit converts real-time vibration information of the facility into a DB and converts the abnormal state occurring in the facility into a DB. In the case of a nozzle for continuous casting, external air inflow, nozzle fastening, clogging of inner diameter, and slag mixing , Steelmaking-casting process facility control and condition analysis system, characterized in that the information on the state of the occurrence of cracks or the dropout of large inclusions is converted into a DB, the correlation analysis information between the operating conditions and the condition is converted into a DB, and information on the use stability standard is converted into a DB.
[Claim 14]
The system of claim 8, wherein the monitoring information calculated through the analysis unit is transmitted using a wired, wireless, or IoT environment.
[Claim 15]
The system of claim 9, further comprising a function of fixing a measurement position by digital image processing of a camera.
[Claim 16]
The method of claim 8, wherein a control signal or information is transmitted in connection with a flow control device and a steelmaking-performing process facility, and in the case of a nozzle for continuous casting, an automatic nozzle position adjustment signal is sent when an abnormality is detected in the nozzle connection. Steelmaking, characterized by sending information about the time point as a warning signal when detecting inclusion dropout, and sending a control signal or replacement signal to the flow control device when the inflow of slag sink vortex and slag is detected, or when the blockage of the inner diameter exceeds a certain value. Control and condition analysis system of performance process equipment.
[Claim 17]
The method of claim 5, wherein the vibration measurement unit continuously measures the vibration data of the steelmaking-performing process facility using a non-contact means, and uses a laser vibration meter, and data mining is used to perform an abnormal state of the process and a nozzle for continuous casting. In the case of, external air inflow, nozzle connection failure, inner diameter clogging, slag mixing, cracking, or large inclusions are diagnosed, and based on the diagnosis contents, control signals are transmitted to peripheral devices for efficient steel making-casting quality through casting process control Steel making-control and condition analysis system of a casting process facility, characterized in that it improves and increases the error rate of molten steel.
[Claim 18]
The method according to any one of claims 1 to 4, wherein the steel-making process equipment is a shroud nozzle or an immersion nozzle.
[Claim 19]
10. The control and condition analysis system for a steel-casting process facility according to any one of claims 5 to 9, wherein the steel-making-casting process facility is a shroud nozzle or an immersion nozzle.