Railway vehicle operation safety diagnosis system and diagnosis method

Abstract

In the present invention, vibration generated by a running vehicle is measured and analyzed by one or more vibration sensors provided to each vehicle of a train, operation information including a running position and a running speed of the vehicle is added, the vibration information is transmitted to a central control system, integrated operation data is analyzed by artificial intelligence learned in advance, and the integrated operation data is transmitted to the central control system. According to the railway vehicle operation safety diagnosis system and method, the abnormal occurrence part of a vehicle or a line is predicted, and artificial intelligence re-learning is achieved through big data analysis.

Description

Technical Field

[0001] The present invention relates to a diagnostic system for real-time diagnosis of abnormalities affecting running safety, such as derailment of railway vehicles, and more particularly to a railway vehicle running safety diagnostic system and method, which measures and analyzes vibrations generated by running vehicles through one or more vibration sensors installed in each vehicle of a train, adds running information including the running position and running speed of the vehicle, and transmits the information to a central control system, analyzes the comprehensive running data through artificial intelligence learned in advance, predicts the abnormal occurrence location of the vehicle or line, and realizes re-learning of artificial intelligence through big data analysis. Background Art

[0002] Railways are track facilities used to transport passengers or materials in large quantities and quickly via railway vehicles. A pair of parallel tracks arranged on the ground guide the travel direction of the wheels arranged opposite to each other along the two sides of the railway vehicle, allowing the railway vehicle to travel in a straight line or in a curve.

[0003] With the help of the friction force formed when the wheels connected to the power source of the railway vehicle come into contact with the track, the wheels roll on the track, allowing the railway vehicle to travel along the track. In order to prevent the wheels of the railway vehicle from detaching from the track during travel, the inner treads of the wheels on both sides in contact with the track protrude outward from the wheel radius to form a flange.

[0004] When a railway vehicle is running, as the wheels move along the upper surface of the track, the railway vehicle travels and the lateral movement direction of the travel direction is controlled by the flange. Therefore, railway vehicles are often in danger of safety accidents caused by derailment. In order to prevent derailment accidents caused by abnormalities of railway vehicles, a diagnostic system that monitors the status of railway vehicles in real time is used to ensure the driving stability of railway vehicles.

[0005] The vehicle diagnostic system used to diagnose abnormal conditions during railway vehicle operation usually follows the condition-based maintenance (CBM) system. In condition-based maintenance, sensors or diagnostic devices are installed at major locations that affect the safety of railway vehicle operation. Detection signals of abnormal conditions occurring during railway vehicle operation measured by the sensors or diagnostic devices are transmitted to the driver or vehicle control system. When an abnormal condition of the vehicle is detected, emergency measures such as deceleration or emergency stop are implemented.

[0006] As an example of such a railway vehicle diagnostic system, Korean Patent Gazette No. 10-2235625 (authorized on March 29, 2021) proposes an abnormal state diagnostic system for a railway vehicle in which, when the wheels of a railway vehicle pass through a diagnostic section, an excitation unit excites the line to generate a sinusoidal wave with a prescribed period on the line, a measuring unit measures the vibration generated in the railway vehicle by the excitation, and the abnormal state of the railway vehicle is diagnosed based on the measured vibration.

[0007] Due to the characteristics of a train consisting of multiple vehicles connected together, derailment may cause a large number of casualties, so it is necessary to continuously monitor the abnormal conditions of railway vehicles during operation. However, the railway vehicle diagnostic system as described above has the problem of being unable to diagnose the abnormal conditions of railway vehicles in sections other than the diagnostic section.

[0008] Furthermore, even if a railway vehicle is in a normal state, a derailment accident may occur if a problem occurs on the line. Existing railway vehicle diagnostic systems that analyze and check abnormal conditions of vehicles using vibration patterns measured by vibration sensors have the problem of being unable to detect abnormalities occurring on the line.

[0009] Prior art literature

[0010] Patent Literature

[0011] Patent Document 1: Korean Patent Gazette No. 10-2235625 (granted on March 29, 2021) Summary of the invention

[0012] Problem that the invention aims to solve

[0013] The object of the present invention is to provide a railway vehicle operation safety diagnosis system and a diagnosis method which can detect abnormal conditions of railway vehicles and lines in real time, distinguish and detect the occurrence of abnormal conditions of railway vehicles or lines, and when an abnormal condition is diagnosed, perform emergency measures such as deceleration or emergency stop on a train including the vehicle or a train running on the line, and inspect and maintain the parts diagnosed with the abnormal condition.

[0014] Means used to solve problems

[0015] The railway vehicle operation safety diagnosis system according to the present invention comprises: a vibration diagnosis module, which measures the vibration generated by the running vehicle through one or more vibration sensors installed in each vehicle of the train, obtains a vibration signal, and analyzes the vibration signal to form vibration data; a vehicle computer, which is installed in each vehicle, receives the vibration data formed in the vibration diagnosis module of the vehicle, and then adds the operation information including the vehicle's running position and running speed to generate operation data; a train comprehensive information device, which summarizes the operation data received from the vehicle computers of the vehicles forming the train, generates comprehensive operation data, and then transmits it to the central control system; a server, which stores the comprehensive operation data transmitted to the central control system to construct big data; a big data analysis module, which analyzes the comprehensive operation data through artificial intelligence learned in advance, predicts the abnormal occurrence location of the vehicle or line, and transmits it to the central control system, realizes re-learning of artificial intelligence from the data obtained by comparing the prediction results with the actual measurement results based on the abnormal occurrence inspection, and the operation data and comprehensive operation data are generated by dividing the entire section of the line where the train runs into sections of specified positions, and the vibration data measured in the vehicle is generated according to the line section unit, and the big data analysis module predicts the abnormal occurrence location of a specific vehicle or line.

[0016] According to an embodiment of the present invention, vibration data is formed by a vibration diagnosis module using an analysis method including one of RMS, vibration level, ride feeling index, and ride feeling level as representative values ​​that can represent the dynamic state of the vehicle.

[0017] According to an embodiment of the present invention, a vibration sensor of a vibration diagnosis module is mounted on a main part of a vehicle including an axle or a dolly.

[0018] According to an embodiment of the present invention, the vibration diagnosis module determines whether the measured vibration data is within the allowable reference range of the vibration value preset in the line section, transmits the determination result together with the vibration data to the vehicle computer, and feeds back the artificial intelligence re-learning result of the big data analysis of the big data analysis module to continuously update the set allowable reference range of the vibration value.

[0019] According to an embodiment of the present invention, the allowable reference range of vibration values ​​is set according to the type of train or vehicle model, the form of the train or vehicle, or the number of vehicle formations, and the comprehensive operation data transmitted from the train comprehensive information device to the central control system includes data according to the type of train or vehicle model, the form of the train or vehicle, or the number of vehicle formations.

[0020] According to an embodiment of the present invention, when abnormal vibration data is measured from multiple trains running in a specific line section, the big data analysis module diagnoses the line section as a dangerous section, or when abnormal vibration data is measured from a specific vehicle running in all line sections, the big data analysis module diagnoses that the vehicle has an abnormality, thereby constructing an abnormality diagnosis map of the running line and the running vehicle to predict the location where the abnormality of the line or vehicle occurs.

[0021] According to an embodiment of the present invention, re-learning of artificial intelligence through a big data analysis module is achieved in the order of data screening process of comprehensive operating data stored in a server, preprocessing process of screened data, conversion process of preprocessed data, data mining process, and pattern analysis process.

[0022] According to an embodiment of the present invention, the pattern analysis process is implemented in a manner selected from linear regression, artificial neural network, K nearest neighbor or unsupervised learning.

[0023] The railway vehicle operation safety diagnosis method according to the present invention comprises: a vibration data forming step, in which a vibration diagnosis module that receives measurement signals from one or more vibration sensors installed in each vehicle of the train measures the vibration generated by the running vehicle, obtains a vibration signal, and analyzes the vibration signal to form vibration data; an operation data generating step, in which a vehicle computer installed in each vehicle receives the vibration data formed in the vibration diagnosis module of the vehicle, and then adds operation information including the vehicle's running position and running speed to generate operation data; a comprehensive operation data generating step, in which the operation data received from the vehicle computers of the vehicles forming the train are summarized by the train comprehensive information device, and after generating the comprehensive operation data, the comprehensive operation data is transmitted to the central control system; an abnormality occurrence prediction step, in which the comprehensive operation data transmitted to the central control system is stored on the server, and the comprehensive operation data is generated by the train comprehensive information device. The system includes analyzing the transmitted comprehensive operation data using artificial intelligence learned in advance in the big data analysis module. When analyzing the comprehensive operation data, the entire section of the line where the train runs is divided into sections of specified positions, and vibration data measured in the vehicle is generated according to the line section unit. When abnormal vibration data is measured from multiple trains running in a specific line section, the line section is diagnosed as a dangerous section, or when abnormal vibration data is measured from a specific vehicle running in all line sections, the vehicle is diagnosed as abnormal, and a prediction result of the line or vehicle abnormality is provided to the central control system; an abnormality occurrence response step, when a line or vehicle abnormality is predicted, the central control system executes response measures based on the abnormality occurrence; a re-learning step, realizing re-learning of the artificial intelligence by comparing the prediction result with the actual measurement result based on the abnormality occurrence inspection.

[0024] According to an embodiment of the present invention, in the vibration data forming step, the vibration diagnosis module determines whether the measured vibration data is within the allowable reference range of the vibration value preset in the line section, and transmits the determination result together with the vibration data to the vehicle computer. In the relearning step, the artificial intelligence relearning result of the big data analysis of the big data analysis module is fed back to the vibration diagnosis module, and the set allowable reference range of the vibration value is continuously updated.

[0025] According to an embodiment of the present invention, in the re-learning step, the re-learning of the artificial intelligence through the big data analysis module is implemented in the order of data screening process of the comprehensive operation data stored in the server, preprocessing process of the screened data, conversion process of the preprocessed data, data mining process, and pattern analysis process. The pattern analysis process is implemented in a method selected from linear regression, artificial neural network, K nearest neighbor or unsupervised learning.

[0026] Effects of the Invention

[0027] According to the present invention, the following effects are achieved: the entire section of the line where the train runs is divided into sections at specified locations according to terrain conditions or necessary conditions for safe operation management and maintenance and repair, section units are generated according to the line, vibration data is individually generated according to the set line section units from the vibration signal obtained by the vibration sensor, and when an abnormality is diagnosed through vibration data analysis, the abnormal state in the vehicle or line can be distinguished for detection, the position of the line where the abnormal state occurs can be determined, and countermeasures to the abnormal state can be quickly implemented.

[0028] According to the present invention, the following effects are achieved: the accuracy of prediction of abnormality occurrence of vehicles or lines by a railway vehicle operation safety diagnosis system can be improved by big data analysis of operation data generated from vibration signals obtained from each vehicle of a train.

[0029] According to the present invention, the following effects are achieved: the data obtained through big data analysis of operating data is used to realize the re-learning of artificial intelligence for predicting abnormalities in vehicles or lines, and the allowable reference range of vibration values ​​obtained through artificial intelligence re-learning is fed back and updated regularly, which can further improve the accuracy of railway vehicle operation safety diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 FIG. 2 is a diagram showing the overall structure of a railway vehicle operation safety diagnosis system according to the present invention.

[0031] Figure 2 The diagram shows a data aggregation process implemented in the process of generating comprehensive operation data from a vibration signal measured by a vibration sensor installed on a vehicle in the railway vehicle operation safety diagnosis system according to the present invention.

[0032] Figure 3 A diagram showing line section units formed by dividing lines in order to distinguish abnormalities occurring in lines or vehicles and determine the line where an abnormal state occurs when an abnormal state of the line is detected in the railway vehicle operation safety diagnosis system according to the present invention.

[0033] Figure 4 The diagram is a diagram showing an example of comparing the vibration data measured by the vibration sensor according to the present invention with the allowable reference range of the vibration value to determine the section position of the track where the abnormality is predicted to occur.

[0034] Figure 5 The diagram is a diagram showing an example of determining a line where an abnormal state occurs based on vibration data measured from a plurality of trains in the railway vehicle operation safety diagnosis system according to the present invention.

[0035] Figure 6 The diagram is a diagram showing an example of specifying a vehicle in an abnormal state based on vibration data measured from a specific vehicle in a running train according to the railway vehicle operation safety diagnosis system of the present invention.

[0036] Figure 7 FIG. 1 is a diagram showing the sequence of the railway vehicle operation safety diagnosis method according to the present invention.

[0037] Description of Reference Numerals

[0038] 1: Central Control System 2: Server

[0039] 10: Train 11: Vehicle

[0040] 12: Train integrated information device 100: Vehicle computer

[0041] 110: Vibration diagnosis module 101: Vibration sensor

[0042] 200: Big Data Analysis Module DETAILED DESCRIPTION

[0043] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0044] The detailed description will be centered around parts necessary for understanding the operation and effects of the present invention.

[0045] When describing the embodiments of the present invention, descriptions of technical contents that are well known in the technical field to which the present invention belongs and are not directly related to the present invention are omitted.

[0046] This is for conveying the gist of the present invention more clearly without ambiguity by omitting unnecessary explanations.

[0047] Furthermore, when describing the components of the present invention, components having the same name may be denoted by different reference numerals depending on the drawings, and components may be denoted by the same reference numerals even in different drawings.

[0048] However, even in this case, it does not mean that the structural element has different functions according to the embodiment, or that it has the same function in different embodiments, and the function of each structural element should be judged based on the description of each structural element in the embodiment.

[0049] Furthermore, unless otherwise specifically defined in this specification, the technical terms used in this specification should be interpreted as the meanings commonly understood by ordinary technicians in the technical field to which the present invention belongs, and should not be interpreted as overly comprehensive or overly narrow meanings.

[0050] Furthermore, unless there is a different meaning in the context, a singular expression used in the present specification includes a plural expression.

[0051] In the present application, terms such as “compose” or “comprise” should not be interpreted as necessarily including all of the multiple structural elements or multiple steps recorded in the specification, and should be interpreted as possibly excluding some of the structural elements or some of the steps, or possibly including additional structural elements or steps.

[0052] In this specification, a vehicle 11 is formed by an axle formed by wheels and axles, an axle box assembly device, a suspension device, a traction device, a braking device, and a bogie as a traveling device formed by a connector, combined with a body carrying passengers or equipped with driving equipment. The train 10 means a railway vehicle in which multiple vehicles 11 are connected in a row by connectors.

[0053] like Figure 1 As shown, the railway vehicle operation safety diagnosis system according to the present invention can generate operation data from the vibration signals obtained from the various vehicles 11 forming the train 10. After being transmitted to the central control system 1, the operation data is analyzed by the big data analysis module 200 to predict the abnormal occurrence location of the vehicle 11 or the line. The operation data stored in the server 2 is analyzed and converted into big data to improve the accuracy of the prediction of the abnormal occurrence of the vehicle 11 or the line through the railway vehicle operation safety diagnosis system.

[0054] like Figure 2 As shown, each vehicle 11 of the train 10 is provided with one or more vibration sensors 101, which measure the vibration transmitted from the line to the vehicle 11 through the wheels or the vibration generated by the vehicle 11 to obtain a vibration signal. The vibration signal obtained by the vibration sensor 101 is transmitted to a vibration diagnosis module 110 installed in each vehicle 11, and the vibration signal measured by the vibration diagnosis module 110 is analyzed to form vibration data.

[0055] The vibration sensor 111 is preferably installed on the axle, trolley and body which are the main parts of each vehicle 11. However, in order to reduce the increase in the construction cost of the railway vehicle operation safety diagnosis system caused by setting up multiple vibration sensors 111, or the overload of the vibration signal analysis and vibration data transmission in the vibration diagnosis module 110, or the overload of the storage capacity of the big data formed by the operation data stored in the server 2, it can also be installed only at a limited position on the axle or trolley.

[0056] Vibration data is formed by analyzing representative values ​​that can represent the dynamic state of the vehicle 11 through the vibration diagnosis module 110. The representative values ​​obtained from the vibration sensor 111 include root mean square (RMS), vibration level, ride feel index, ride feel level, etc. The vibration data is formed by an analysis method selected therefrom.

[0057] RMS represents the effective output that serves as a benchmark for the average output of the vibration signal continuously output by the vibration sensor 111. After taking the arithmetic average of the squares of the vibration signal output values, the vibration level is calculated by taking the square root of the arithmetic average, and the intensity of the vibration signal measured by the vibration sensor 111 is converted into decibels (dB) as an absolute value.

[0058] The ride feel index is derived by passing the vibration acceleration signal caused by the linear vibration in the front-to-back direction, the left-to-right direction, and the up-and-down direction of the vehicle 11 measured by the vibration sensor 111 through a weighting filter, obtaining the effective value at each specified interval, obtaining the cumulative frequency distribution, and substituting the vibration acceleration equivalent to 95% of the cumulative frequency distribution into the following ride feel index formula.

[0059] Ride experience index

[0060] At this time, a XP95 This means the vibration acceleration equivalent to 95% of the cumulative frequency distribution of the front-to-back vibration, a YP95 This means the vibration acceleration equivalent to 95% of the cumulative frequency distribution of left-right vibration, a ZP95 This means the vibration acceleration corresponding to 95% of the cumulative frequency distribution of vertical vibration.

[0061] The ride quality level is derived by applying a ride quality filter equivalent to the human perception evaluation sensitivity to the effective value of the body vibration acceleration of the vehicle 11 measured by the vibration sensor 111 in accordance with ISO 2631 and expressing it in decibel (dB) units. The result is then substituted into the following ride quality level formula.

[0062] Ride experience level

[0063] At this time, a w Indicates the effective value of the vibration acceleration to which the ride filter is applied, a ref Indicates the reference value of vibration to which the ride filter is applied.

[0064] Furthermore, the vibration diagnosis module 110 compares the vibration data formed in the above-described manner with an allowable reference range of vibration values ​​preset in the vibration diagnosis module 110 to determine whether the vibration data is within the allowable reference range or exceeds the allowable reference range.

[0065] The line is set in various terrain environments such as flat ground, tunnels, or bridges, so the vibration data obtained by the vibration diagnosis module 110 can be formed differently in each section of the line. When the allowable reference range of the vibration value preset in the vibration diagnosis module 110 is uniformly set, according to the terrain environment in which the line is set, the vibration data obtained from the line or vehicle 11 in a normal state may be misjudged as an abnormal state, or conversely, the vibration data obtained from the line or vehicle 11 in an abnormal state may be misjudged as a normal state.

[0066] Therefore, if Figure 3 As shown, the entire section of the line on which the train 10 runs is divided into sections of specified positions according to the terrain conditions or the necessary conditions for safe operation management and maintenance, and section units are generated according to the line. Thus, the vibration signal measured by the vibration sensor 111 and the vibration data obtained by the vibration diagnosis module 110 are also individually generated according to the line section units.

[0067] Furthermore, the allowable reference range of the vibration value preset in the vibration diagnosis module 110 is also individually set according to the route section unit.

[0068] Furthermore, as the type of vehicle 11 changes, such as the model of the vehicle 11 such as a locomotive, a passenger car or a freight car, the form of the train 10 or vehicle 11 formed by a specific model, the number of vehicles 11 in a formation, the method of connecting the vehicles 11, the change in the vibration mode of the vehicle 11 caused by the use of diesel or electricity, the vibration signal measured by the vibration sensor 111 and the vibration data obtained in the vibration diagnosis module 110 also change. Therefore, the allowable reference range of the vibration value preset in the vibration diagnosis module 110 is individually set according to multiple types including the model of the train 10 or vehicle 11, the form of the train 10 or vehicle 11 or the number of vehicles 11 in a formation.

[0069] The vibration data generated in the vibration diagnosis module 110 is transmitted to the vehicle computer 100 installed in each vehicle 11. After each vehicle computer 100 receives the vibration data generated in the vibration diagnosis module 110 of the vehicle 11, it adds the operating information of the driving position and driving speed of the vehicle 11 obtained from the global positioning system (GPS) data transmitted through the train comprehensive information device (TCMS, Train Control and Monitoring System) 12 to generate operating data.

[0070] By adding the operation information to the operation data, the position of the line section from which the vibration data included in the operation data is obtained can be determined, whereby when an abnormal state of the line is determined, emergency measures can be quickly taken to perform inspection or maintenance work on the line.

[0071] The operation data generated by each vehicle computer 100 installed in each vehicle 11 forming the train 10 is transmitted to the train comprehensive information device 12. The train comprehensive information device 12 summarizes the received operation data, generates comprehensive operation data, and transmits it to the central control system 1.

[0072] The comprehensive operation data transmitted from the train comprehensive information device 12 of each train 10 is transmitted to the server 2 connected to the central control system 1 via the network and stored, and the big data of the railway vehicle operation safety diagnosis system is constructed from the comprehensive operation data accumulated and stored in the server 2.

[0073] The big data analysis module 200 is connected to the central control system 1 and the server 2 through the network, receives the comprehensive operation data transmitted from the train comprehensive information device 12 of each train 10 in real time, and performs analysis of the comprehensive operation data through pre-learned artificial intelligence.

[0074] According to the analysis result of the big data analysis module 200 on the comprehensive operation data, Figure 4 As shown, when the vibration data measurement value exceeds the allowable reference range of the vibration value and an abnormality is predicted to occur in the vehicle 11 or the track, the section position of the vehicle 11 or the track of the train 10 where the abnormality is predicted to occur is determined, and the prediction result is transmitted to the central control system 1.

[0075] The central control system 1 that receives the prediction results executes emergency measures such as deceleration or emergency stop on the train 10 connected to the vehicle 11 where the abnormality is predicted to occur, or executes emergency measures such as deceleration or emergency stop on the train 10 close to the line section where the abnormality is predicted to occur, according to the degree of the symptoms of the prediction results.

[0076] When deceleration measures are implemented, after the railway operation business hours end, personnel are deployed to the line section or vehicle 11 where the abnormality is predicted to occur for inspection and maintenance work. When emergency stop measures are implemented, personnel are immediately deployed to the line section or vehicle 11 for inspection and maintenance work.

[0077] In addition, the big data analysis module 200 compares the prediction results of the vehicle 11 or line abnormality and the actual measured results of the abnormality inspection after the central control system 1 executes emergency measures, obtains big data on the accuracy of the abnormality prediction analysis, and realizes re-learning of artificial intelligence from the obtained data.

[0078] Based on the artificial intelligence re-learning results of the big data analysis through the big data analysis module 200, the allowable reference range of the vibration value obtained is fed back to the allowable reference range of the vibration value set in the vibration diagnosis module 110 and the big data analysis module 200, and the existing set allowable reference range of the vibration value of the vibration diagnosis module 110 and the big data analysis module 200 is updated.

[0079] In each prescribed cycle set by the administrator of the railway vehicle operation safety diagnosis system, artificial intelligence re-learning is performed by regularly accumulating big data analysis of the comprehensive operation data stored on the server 2, continuously updating the allowable reference range of the vibration value, and improving the accuracy of the prediction results of the vehicle 11 or line abnormality occurrence of the big data analysis module 200.

[0080] The re-learning of the artificial intelligence through the big data analysis module 200 is implemented in the order of the process of screening the comprehensive operation data stored in the server 2, the preprocessing process of preprocessing the screened data, the conversion process of converting the preprocessed data, the data mining process and the pattern analysis process.

[0081] In the pattern analysis process, the pattern analysis is implemented in a manner selected from linear regression, artificial neural network, K-nearest neighbor or unsupervised learning.

[0082] Linear regression is an analysis technique that uses a least squares method or other technique to model the linear correlation (regression equation) of known data values ​​such as vibration data or comprehensive operating data measured by the vibration sensor 111, and then infers unknown data and parameters from the modeled linear correlation.

[0083] Artificial neural networks are machine learning models that mimic the structure of human neurons through software. They repeatedly learn in the direction that is expected to minimize the error with previously learned data, and update weight values ​​and thresholds to achieve artificial intelligence.

[0084] The K nearest neighbor algorithm is an algorithm that uses the Euclidean distance calculation method to refer to the labels of other data that are close to the data for classification. It finds K data adjacent to the data to be identified and classifies the data into multiple categories based on the label of the data. The algorithm is simple, the training phase is fast, and it shows excellent performance in numerical-based data classification work.

[0085] The K nearest neighbor algorithm is formed by a distance-based model. Therefore, in order to facilitate the explanation of the differences between various variables, it is necessary to apply normalization such as minimum-maximum normalization or z-score normalization to readjust the variable values ​​to the standard range. In order to classify data, the number of K is preferably set to an odd number.

[0086] Unsupervised learning is an analytical technique that uses machine learning algorithms to analyze unlabeled data sets without explicit programming and by providing a large amount of data, allowing the system to learn and improve autonomously, grouping data with the same or similar characteristics and classifying them into clusters.

[0087] At this time, the allowable reference range of the vibration value is set according to multiple types of distinctions, such as the vehicle type of the train 10 or the vehicle 11 such as KTX or SRT or their detailed model classifications, the form type of the train 10 or the vehicle 11 such as a power-concentrated KTX or a power-dispersed KTX joint, and the marshaling number type of the vehicle 11 such as a 10-unit marshaling KTX or a 20-unit marshaling KTX. When the occurrence of a vehicle 11 or line abnormality is predicted by comparing the vibration data and the allowable reference range of the vibration value in the vibration diagnosis module 110 or by comparing the comprehensive operation data and the allowable reference range of the vibration value in the big data analysis module 200, the allowable reference range of the vibration value set according to the type distinction is applied to minimize the error of the prediction result of the occurrence of the vehicle 11 or line abnormality.

[0088] Furthermore, various sensors are used to confirm the status of the main devices of the vehicle in real time. Unlike the existing vehicle diagnostic system that uses a condition-based maintenance (CBM) method that collects and analyzes fault records or maintenance operation data, the railway vehicle operation safety diagnostic system of the present invention can diagnose abnormalities in the vehicle 11 and predict abnormalities in the line. The process of distinguishing between abnormalities in the vehicle 11 and abnormalities in the line based on vibration data measured by the same vibration sensor 111 and making predictions is as follows.

[0089] like Figure 5As shown, when the vibration data measured from multiple trains 10 running in a specific line section exceeds the allowable reference range of vibration values ​​based on the line section and the type of vehicle 11 of the train 10 and is determined to have generated abnormal vibration data, the line section is diagnosed as a dangerous section. In order to prevent errors in determining the line dangerous section, whether the line is abnormal is determined based on the vibration data measured in more than three trains 10.

[0090] And, if Figure 6 As shown, vibration data measured from a specific vehicle 11 of a train 10 running along all line sections periodically or intermittently exceeds an allowable reference range of vibration values ​​according to the line section and the type of vehicle 11 of the train 10 in multiple line sections, and is determined to generate abnormal vibration data. When abnormal vibration data of the same or similar pattern is not generated in the vehicles 11 of other trains 10, it is diagnosed that the vehicle 11 in which the abnormal vibration data was measured is abnormal.

[0091] In order to distinguish the occurrence of abnormalities in vehicles 11 or lines and make predictions, the big data analysis module 200 constructs an abnormality diagnosis map of the operating line and the operating vehicle 11 through the above-mentioned determination method, distinguishes the occurrence of abnormalities in the line or vehicle 11 based on the abnormality diagnosis map and makes diagnoses, and accurately determines the line section or vehicle 11 where the abnormal state occurs.

[0092] The railway vehicle operation safety diagnosis method using the railway vehicle operation safety diagnosis system constructed as above is as follows Figure 7 Follow the steps shown.

[0093] In the vibration data forming step S10 , the vibration diagnosis module 110 receives a measurement signal from a vibration sensor 111 installed in each vehicle 11 of the train 10 , obtains a vibration signal generated by the running vehicle 11 , and forms vibration data by analyzing the vibration signal.

[0094] In the operation data generating step S20, the vibration data generated in the vibration data forming step S10 is received by a vehicle computer 100 installed in the same vehicle 11 as the vehicle 11 for obtaining the vibration data. When the vibration data is obtained, operation information including the driving position which is the position of the route section in which the vehicle 11 is traveling and the driving speed which is the speed at which the vehicle 11 passes through the route section is attached to the vibration data to generate the operation data.

[0095] In the comprehensive operation data generation step S30, the operation data received from the vehicle computers 100 of the various vehicles 11 forming the train 10 in the operation data generation step S20 are summarized by the train comprehensive information device 12 to generate comprehensive operation data, and the train comprehensive information device 12 transmits the generated comprehensive operation data to the central control system 1.

[0096] In the abnormality occurrence prediction step S40, the comprehensive operation data transmitted from the train comprehensive information device 12 to the central control system 1 is stored on the server 2, and the transmitted comprehensive operation data is analyzed in real time by the pre-learned artificial intelligence included in the big data analysis module 200. After diagnosing the abnormality of the line or vehicle 11, the prediction result of the abnormality of the line or vehicle 11 is provided to the central control system 1.

[0097] In the abnormality occurrence response step S50, when an abnormality is predicted to occur in the line or vehicle 11 based on the abnormality occurrence result predicted in the abnormality occurrence prediction step S40, the central control system 1 executes response measures including deceleration or emergency stop of the train 10 running on the line or the train 10 including the vehicle 11 and the deployment of personnel for inspection and maintenance work according to the type of abnormality occurrence or the degree of the abnormality occurrence symptoms. When the inspection of the line or vehicle 11 is completed in response to the abnormality occurrence, the actual measurement results generated based on the inspection are stored on the server 2.

[0098] In the relearning step S60, artificial intelligence relearning is achieved by comparing the predicted results with the actual measured results generated according to the inspection results of the line or vehicle 11. After the relearning is completed, the optimized allowable reference range of the vibration value is transmitted to the vibration diagnosis module 110 and the big data analysis module 200 to update the existing allowable reference range of the vibration value.

[0099] In the relearning step S60, the data screening process (step S61) of screening the comprehensive operation data stored in the server 2, the preprocessing process of preprocessing the screened data (step S62), the conversion process of the preprocessed data (step S63), the data mining process (step S64) and the pattern analysis process (step S65) are carried out in this order.

[0100] The optimization of the allowable reference range of vibration values ​​executed according to the re-learning step S60 and the updating of the allowable reference range of existing vibration values ​​of the vibration diagnosis module 110 and the big data analysis module 200 are automatically executed in each cycle set by the railway vehicle operation safety diagnosis system administrator. By continuously updating the allowable reference range of vibration values, the accuracy of the prediction results of vehicle 11 or line abnormality is improved, and the occurrence rate of safety accidents such as derailment of train 11 is reduced.

[0101] The embodiments of the present invention are described with reference to the above contents, but a person skilled in the art in the art to which the present invention belongs should understand that the present invention can be implemented in other specific forms without changing its technical concept or essential features.

[0102] Therefore, the embodiments described above should be understood to be illustrative in all aspects and not restrictive, and the scope of the present invention described in the above detailed description is expressed according to the scope of the patent claims described later, and should be interpreted as the meaning and scope of the patent claims and all changes or deformations derived from their equivalent concepts are included in the scope of the present invention.

Claims

1. A railway vehicle operation safety diagnostic system, characterized in that: include: A vibration diagnosis module (110) measures vibrations generated by a running vehicle (11) through one or more vibration sensors (101) provided on each vehicle (11) of a train (10), obtains a vibration signal, and analyzes the vibration signal to form vibration data; A vehicle computer (100) is installed in each vehicle (11), receives vibration data generated in a vibration diagnosis module (110) of the vehicle (11), and then adds operation information including a driving position and a driving speed of the vehicle (11) to generate operation data; The train integrated information device (12) collects the operation data received from the vehicle computers (100) of the vehicles (11) forming the train (10), generates integrated operation data, and transmits it to the central control system (1); The server (2) stores the comprehensive operation data transmitted to the central control system (1) to construct big data; as well as The big data analysis module (200) analyzes the comprehensive operation data through the artificial intelligence learned in advance, predicts the abnormality occurrence position of the vehicle (11) or the line, transmits it to the central control system (1), and realizes the re-learning of the artificial intelligence from the data obtained by comparing the prediction result with the actual measurement result based on the abnormality occurrence inspection. The operation data and the comprehensive operation data are generated by dividing the entire section of the line on which the train (10) runs into sections of specified positions, and the vibration data measured in the vehicle (11) is generated according to the line section unit, and the abnormal occurrence location of a specific vehicle (11) or line is predicted by a big data analysis module (200).

2. The railway vehicle operation safety diagnostic system according to claim 1, characterized in that: Vibration data is generated by a vibration diagnosis module (110) using an analysis method including one of a root mean square (RMS), a vibration level, a ride feeling index, and a ride feeling level as representative values ​​that can represent the dynamic state of the vehicle (11).

3. The railway vehicle operation safety diagnostic system according to claim 1, characterized in that: A vibration sensor (111) of a vibration diagnosis module (110) is mounted on a main part of a vehicle (11) including an axle or a trolley.

4. The railway vehicle operation safety diagnostic system according to claim 1, characterized in that: The vibration diagnosis module (110) determines whether the measured vibration data is within the allowable reference range of the vibration value preset in the route section, transmits the determination result together with the vibration data to the vehicle computer (100), and feeds back the artificial intelligence re-learning result of the big data analysis of the big data analysis module (200), so as to continuously update the set allowable reference range of the vibration value.

5. The railway vehicle operation safety diagnostic system according to claim 4, characterized in that: The allowable reference range of the vibration value is set according to the type of train (10) or vehicle (11), the form of the train (10) or vehicle (11), or the number of vehicles (11). The integrated operation data transmitted from the train integrated information device (12) to the central control system (1) includes data according to the type of vehicle model of the train (10) or vehicle (11), the form of the train (10) or vehicle (11), or the number of vehicles (11).

6. The railway vehicle operation safety diagnostic system according to claim 1, characterized in that: When abnormal vibration data is measured from a plurality of trains (10) running in a specific line section, the big data analysis module (200) diagnoses the line section as a dangerous section, or When abnormal vibration data is measured from a specific vehicle (11) operating in all line sections, the big data analysis module (200) diagnoses that the vehicle (11) has an abnormality. Thus, an abnormality diagnosis map of the running line and the running vehicle (11) is constructed to predict the abnormality occurrence location of the line or the vehicle (11).

7. The railway vehicle operation safety diagnostic system according to claim 1, characterized in that: Re-learning of artificial intelligence through the big data analysis module (200) is achieved in the order of data screening process of the comprehensive operation data stored in the server (2), preprocessing process of the screened data, conversion process of the preprocessed data, data mining process, and pattern analysis process.

8. The railway vehicle operation safety diagnosis system according to claim 7, characterized in that: The pattern analysis process is implemented in a manner selected from linear regression, artificial neural network, K nearest neighbor or unsupervised learning.

9. A railway vehicle operation safety diagnosis method, characterized in that: include: A vibration data forming step (S10), wherein a vibration diagnosis module (110) receives a measurement signal from one or more vibration sensors (111) provided on each vehicle (11) of the train (10), measures the vibration generated by the vehicle (11) in operation, obtains a vibration signal, and analyzes the vibration signal to form vibration data; An operation data generating step (S20) is to generate operation data by adding operation information including a driving position and a driving speed of the vehicle (11) after receiving vibration data generated in a vibration diagnosis module (110) of the vehicle (11) through a vehicle computer (100) installed in each vehicle (11); a comprehensive operation data generating step (S30) of collecting the operation data received from the vehicle computers (100) of the vehicles (11) forming the train (10) through the train comprehensive information system (TCMS) 12, generating the comprehensive operation data, and transmitting the generated comprehensive operation data to the central control system (1); The abnormality occurrence prediction step (S40) stores the integrated operation data transmitted to the central control system (1) on the server (2), analyzes the transmitted integrated operation data through artificial intelligence learned in advance in the big data analysis module (200), divides the entire section of the line where the train (10) runs into sections of specified positions when analyzing the integrated operation data, and generates vibration data measured in the vehicle (11) according to the line section unit. When abnormal vibration data is measured from multiple trains (10) running in a specific line section, the line section is diagnosed as a dangerous section, or when abnormal vibration data is measured from a specific vehicle (11) running in all line sections, the vehicle (11) is diagnosed as abnormal, and the prediction result of the line or vehicle (11) abnormality occurrence is provided to the central control system (1); an abnormality occurrence response step (S50), when it is predicted that an abnormality occurs in a line or a vehicle (11), the central control system (1) executes response measures according to the abnormality occurrence; as well as The relearning step (S60) realizes the relearning of artificial intelligence by comparing the predicted results with the actual measured results based on the abnormality inspection.

10. The railway vehicle operation safety diagnosis method according to claim 9, characterized in that: In the vibration data forming step (S10), the vibration diagnosis module (110) determines whether the measured vibration data is within the allowable reference range of the vibration value preset in the route section, and transmits the determination result together with the vibration data to the vehicle computer (100). In the relearning step (S60), the artificial intelligence relearning result of the big data analysis by the big data analysis module (200) is fed back to the vibration diagnosis module (110), and the allowable reference range of the set vibration value is continuously updated.

11. The railway vehicle operation safety diagnosis method according to claim 9, characterized in that: In the relearning step (S60), the relearning of the artificial intelligence by the big data analysis module (200) is carried out in the order of a data screening process (S61) of the comprehensive operation data stored in the server (2), a preprocessing process (S62) of the screened data, a conversion process (S63) of the preprocessed data, a data mining process (S64), and a pattern analysis process (S65). The pattern analysis process (S65) is implemented in a manner selected from linear regression, artificial neural network, K nearest neighbor or unsupervised learning.

Images