Fault diagnosis method, system and equipment based on turnout current characteristics and medium

Through the classification, cleaning and clustering analysis of action current data, the problem of low timeliness and accuracy of turntable current analysis is solved, more accurate and timely fault diagnosis is achieved, and the reliability and operational efficiency of turntable equipment are improved.

CN118940177APending Publication Date: 2024-11-12HEFEI GOCOM INFORMATION &TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410994782.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

When the prior art analyzes the switch status based on the switch current, the timeliness and accuracy are low, resulting in insufficient timely and accurate fault diagnosis, which may affect train safety and transportation efficiency.

Method used

By obtaining the action current data set, classifying it into abnormal and normal data, data cleaning is performed, current curves are generated and divided into interval segments, multi-dimensional data features are extracted, and turntable state is determined by cluster analysis to improve the accuracy and timeliness of the analysis.

Benefits of technology

It improves the timeliness and accuracy of turntable status analysis, can promptly detect abnormal situations of equipment, predict possible failures in advance, improve the reliability and availability of equipment, and reduce maintenance costs and downtime.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940177A_ABST
    Figure CN118940177A_ABST
Patent Text Reader

Abstract

The invention relates to industrial railway diagnosis and prediction, and discloses a turnout current characteristic-based fault diagnosis method, which comprises the following steps of: acquiring an action current data set of target turnout equipment within a preset time, classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and performing fault diagnosis on the normal data; a turnout current data set is generated according to the cleaned data and normal data, a current curve is generated according to the turnout current data set, and the current curve is divided into a plurality of preset interval sections according to a first-order derivative of the current curve; and collecting the multi-dimensional data features of the current curve in each interval section into a data feature set of the current curve, performing clustering calculation on the data feature set by using a preset target function, and determining the state of the turnout where the target turnout equipment is located according to a calculation result. According to the invention, the timeliness and accuracy of analyzing the turnout state according to the turnout current can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of industrial railway diagnosis and prediction, and in particular to a fault diagnosis method, system, equipment and medium based on turnout current characteristics. Background Art

[0002] The turnout switching system is a key railway equipment used to adjust the direction of train travel, which is directly related to driving safety and transportation efficiency. With the rapid development of my country's railway transportation, including high-speed rail and urban rail transit systems, the speed and frequency of trains are constantly increasing, which makes the stable operation of turnout equipment particularly critical. Failure of turnout equipment may cause train delays or even accidents, and in serious cases may cause casualties and property losses.

[0003] Therefore, it is imperative to effectively monitor the status, diagnose and predict faults of turnout equipment. Modern monitoring systems can collect real-time turnout operation data, including position, electrical signals, mechanical operation status, etc. However, the current existing technology has the problem of low timeliness and accuracy in analyzing the turnout status based on turnout current. Summary of the invention

[0004] The present invention provides a fault diagnosis method, system, electronic device and computer-readable storage medium based on turnout current characteristics, the main purpose of which is to solve the problem of low timeliness and accuracy in analyzing the turnout status based on the turnout current.

[0005] To achieve the above object, the present invention provides a fault diagnosis method based on turnout current characteristics, comprising:

[0006] Obtaining an action current data set of a target turnout device within a preset time;

[0007] Classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and generating a turnout current data set based on the cleaned data and the normal data;

[0008] generating a current curve according to the turnout current data set, and dividing the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve;

[0009] Aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve;

[0010] The data feature set is clustered using a preset objective function, and the state of the turnout where the target turnout device is located is determined based on the calculation result.

[0011] In one embodiment of the present invention, the classifying the action current data set into abnormal data and normal data includes:

[0012] Calculating a standard deviation of the action current data set;

[0013] The data in the action current data set that is larger than the standard deviation is regarded as abnormal data;

[0014] The data in the action current data set that is less than or equal to the standard deviation is regarded as normal data.

[0015] In one embodiment of the present invention, the step of cleaning the abnormal data includes:

[0016] Filling the null value data in the abnormal data by using an interpolation algorithm;

[0017] The non-null data in the abnormal data are replaced by the current mean value in the normal data.

[0018] In one embodiment of the present invention, dividing the current curve into a plurality of preset interval segments according to the first-order derivative of the current curve includes:

[0019] Obtaining a state conversion timestamp of the turnout where the turnout device is located, and removing corresponding data points on the current curve according to the state conversion timestamp to obtain an optimized curve;

[0020] Calculating the first-order derivative of each point on the optimization curve;

[0021] Selecting, in chronological order, a first data point and a second data point on the optimization curve whose first-order derivative is less than a preset derivative threshold;

[0022] dividing the curve before the first data point on the current curve into the first stage of the current curve;

[0023] dividing a curve between the first data point and the second data point on the current curve into a second stage of the current curve;

[0024] The portion of the current curve after the second data point is divided into a third stage of the current curve.

[0025] In one embodiment of the present invention, the step of aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve includes:

[0026] taking the maximum value, the waveform factor and the kurtosis calculated according to the current curve in the first stage and the third stage as the first data feature;

[0027] The average value, standard deviation, root mean square, maximum value, extreme value, waveform factor, pulse factor and mode calculated according to the current curve in the second stage are used as second data features;

[0028] The first data feature and the second data feature are aggregated into a data feature set of the current curve.

[0029] In one embodiment of the present invention, the clustering calculation of the data feature set using a preset objective function includes:

[0030] Converting the data feature set into a numerical matrix;

[0031] Initializing the numerical matrix using random numbers;

[0032] Calculate the cluster center of each data in the data feature set;

[0033] Using the cluster centers to update the initialized numerical matrix to obtain an updated matrix;

[0034] Iteratively calculating the update matrix using a preset objective function to obtain a membership matrix;

[0035] The clustering result of each data point in the data feature set is determined according to the membership matrix.

[0036] In one embodiment of the present invention, determining the state of the turnout where the target turnout device is located according to the calculation result includes:

[0037] Selecting target data one by one from the data feature set, and obtaining reference data points in the cluster corresponding to the target data;

[0038] Calculating a first average distance between the target data and the reference data points;

[0039] Determine the adjacent clusters of the cluster corresponding to the target data, calculate the average distance between the target data and all data points in the adjacent clusters, and obtain a second average distance;

[0040] Calculate the silhouette coefficient of the target data in the membership matrix one by one according to the first average distance and the second average distance;

[0041] Calculate an average silhouette coefficient based on the silhouette coefficient of the target data;

[0042] If the average contour coefficient is greater than a preset contour coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a healthy state;

[0043] If the average profile coefficient is equal to the preset profile coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a sub-healthy state;

[0044] If the average contour coefficient is less than a preset contour coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a fault state.

[0045] In order to solve the above problems, the present invention also provides a fault diagnosis system based on turnout current characteristics, the system comprising:

[0046] A data acquisition module, used to acquire an action current data set of a target turnout device within a preset time;

[0047] A data cleaning module, used for classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and generating a turnout current data set according to the cleaned data and the normal data;

[0048] A region division module, used for generating a current curve according to the turnout current data set, and dividing the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve;

[0049] A feature extraction module, used for aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve;

[0050] The data analysis module is used to perform clustering calculation on the data feature set using a preset objective function, and determine the state of the turnout where the target turnout equipment is located according to the calculation result.

[0051] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0052] at least one processor; and,

[0053] a memory communicatively connected to the at least one processor; wherein,

[0054] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the steps of the above-mentioned method for fault diagnosis based on turnout current characteristics.

[0055] In order to solve the above problem, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the steps of the above-mentioned method for diagnosing faults based on turnout current characteristics.

[0056] In the present invention, cleaning the abnormal data in the action current data set can remove the abnormal current value, improve the accuracy of the action current data set, and avoid analysis errors caused by erroneous data. The cleaned data and the original data are aggregated into a turnout current data set for comprehensive data analysis to improve data analysis efficiency; dividing the current curve into multiple preset intervals according to the first-order derivative is helpful for more detailed analysis and understanding of the characteristics of current changes, and is helpful for identifying and locating specific events or anomalies, thereby achieving more accurate fault diagnosis. Aggregating the multidimensional data features of the current curve in each interval into a data feature set is helpful for in-depth analysis and comparison of the behaviors of different intervals, and is helpful for identifying the changing trend of the current characteristics in each interval, making the data features more targeted; cluster analysis classifies turnout equipment in similar states into the same cluster, which is helpful for identifying the general working state or potential problems of the equipment, and can timely discover the abnormal situation of the equipment, predict possible faults in advance, and improve the reliability and availability of the equipment. Therefore, a fault diagnosis method, system, electronic device and computer-readable storage medium based on turnout current characteristics proposed in the present invention mainly aims to improve the timeliness and accuracy of analyzing the turnout state based on the turnout current. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A schematic flow chart of a method for diagnosing a fault based on turnout current characteristics provided by one embodiment of the present invention;

[0058] Figure 2 A functional module diagram of a fault diagnosis system based on turnout current characteristics provided by one embodiment of the present invention;

[0059] Figure 3 A schematic diagram of the structure of an electronic device for implementing a method for diagnosing faults based on turnout current characteristics provided in one embodiment of the present invention.

[0060] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0062] The embodiment of the present application provides a method for diagnosing faults based on turnout current characteristics. The execution subject of the method for diagnosing faults based on turnout current characteristics includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for diagnosing faults based on turnout current characteristics can be executed by software or hardware installed in a terminal device or a server device.

[0063] Reference Figure 1 FIG. 1 is a flow chart of a method for diagnosing a fault based on a turnout current characteristic provided by an embodiment of the present invention. In this embodiment, the method for diagnosing a fault based on a turnout current characteristic includes:

[0064] S1. Obtain the action current data set of the target turnout equipment within a preset time.

[0065] In the embodiment of the present invention, the target turnout equipment includes two switch machines for synthesizing the current curve. The data server collects the action current data within a preset time, which can include the complete life cycle of the turnout equipment, that is, the whole process data from the turnout equipment being normal, to the turnout equipment being deteriorated, and finally to the turnout equipment being faulty, and the turnout equipment being restored to normal after maintenance.

[0066] The benefit of obtaining the action current data set of the target turnout equipment within a preset time is that it can monitor the equipment status in real time, quickly diagnose faults, optimize maintenance plans, improve equipment performance, and provide data-driven decision support, thereby significantly improving the reliability, efficiency and operation management of the equipment.

[0067] S2. Classify the action current data set into abnormal data and normal data, clean the abnormal data, and generate a turnout current data set based on the cleaned data and the normal data.

[0068] In an embodiment of the present invention, the classification of the action current data set into abnormal data and normal data includes: calculating the standard deviation of the action current data set; treating the data in the action current data set that is greater than the standard deviation as abnormal data; and treating the data in the action current data set that is less than or equal to the standard deviation as normal data.

[0069] By classifying the action current dataset into abnormal data and normal data through standard deviation, data points that are beyond the normal range can be effectively identified and eliminated, thereby improving the quality and reliability of the action current dataset, helping to ensure the accuracy of analysis and model building, while reducing errors caused by abnormal data, making predictions based on the action current dataset more reliable and accurate.

[0070] In the embodiment of the present invention, the data cleaning of the abnormal data includes: filling the null value data in the abnormal data with an interpolation algorithm; and replacing the non-null data in the abnormal data with the current mean value in the normal data.

[0071] In the embodiment of the present invention, the null-value data in the abnormal data is linearly interpolated using the adjacent current data of the null-value data.

[0072] Specifically, cleaning abnormal data in the current data set can repair abnormal current values ​​and interpolate and fill in null value data, thereby improving the quality of the action current data set, ensuring that the action current data set is more accurate, and avoiding analysis errors caused by erroneous data.

[0073] The cleaned data and the uncleaned data are aggregated into a turnout current data set, making the analysis process simpler and more consistent. Comprehensive data exploration and analysis can be performed in the turnout current data set, thereby improving the analysis accuracy of the turnout current data set.

[0074] S3. Generate a current curve according to the turnout current data set, and divide the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve.

[0075] In the embodiment of the present invention, the current curve of the turnout current data set may be generated by two switch machines in the turnout equipment.

[0076] In an embodiment of the present invention, the current curve is divided into a plurality of preset interval segments according to the first-order derivative of the current curve, including: obtaining a state transition timestamp of the turnout where the turnout equipment is located, and eliminating corresponding data points on the current curve according to the state transition timestamp to obtain an optimized curve; calculating the first-order derivative of each point on the optimized curve; selecting, in chronological order, a first data point and a second data point on the optimized curve whose first-order derivative is less than a preset derivative threshold; dividing the curve before the first data point on the current curve into the first stage of the current curve; dividing the curve between the first data point and the second data point on the current curve into the second stage of the current curve; and dividing the curve after the second data point on the current curve into the third stage of the current curve.

[0077] In detail, the switching state, that is, the display state of the turnout during the rotation process, includes a positioning state, a turnout unlimited state, a reverse position process state, etc.

[0078] In detail, the first data point whose first-order derivative is smaller than a preset derivative threshold value on the optimization curve is selected in chronological order as the first data point, and the last data point whose first-order derivative is smaller than the preset derivative threshold value is selected as the second data point.

[0079] For example, the first-order derivative results of the optimization curve corresponding to the current data from the first second to the fifth second are 0.5, 0.08, 0.9, 0.06, and 0.03, and the preset derivative threshold is 0.1. The first derivative result 0.08 that is less than the preset derivative threshold 0.1 corresponds to the second second current data, so the second second current data is used as the first data point, and the last derivative result 0.06 that is less than the preset derivative threshold 0.1 corresponds to the fourth second current data, so the fourth second current data is used as the second data point. Therefore, the first stage includes the optimization curves corresponding to the first-order derivative results of 0.5 and 0.08, the second stage includes the optimization curves corresponding to the first-order derivative results of 0.9, and the third stage includes the optimization curves corresponding to the first-order derivative results of 0.06 and 0.03.

[0080] The present invention divides the current curve into a plurality of preset interval segments by taking the first-order derivative of the current curve, which helps to more finely analyze and understand the characteristics of current changes, and can help identify and locate specific events or anomalies, such as mutations, peaks or periodic changes, thereby providing a basis for more accurate fault diagnosis, event detection or system optimization.

[0081] S4. Aggregate the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve.

[0082] In an embodiment of the present invention, the multidimensional data features include mean value, standard deviation, root mean square, maximum value, difference between maximum and minimum values, waveform factor, impulse factor, skewness, kurtosis, mode, etc.

[0083] Specifically, the data features extracted from each interval segment are different. The data features extracted in the first stage are maximum value, waveform factor and kurtosis. The reasons are as follows: the first stage is in the rising period, the number of data is small, and the features of mean value, standard deviation, root mean square and mode are not obvious; because the minimum value is 0, the difference between the maximum and minimum values ​​is equal to the maximum value; the pulse factor represents the ratio of the peak value to the average value, and due to the small number of data, it is approximately proportional to the maximum value; the skewness represents the statistic of the degree of skewness of the data distribution. There is no symmetry in the data in the first stage, so it is meaningless to extract the skewness feature.

[0084] The reason for selecting features in the third stage is the same as that in the first stage, so the data features extracted in the third stage are also maximum value, waveform factor and kurtosis.

[0085] The data features extracted in the second stage are mean, standard deviation, root mean square, maximum, difference between maximum and minimum, waveform factor, impulse factor and mode. The reasons are as follows: the data in the transition state in the second stage is basically in a stable state. Skewness is a statistic that describes the direction and degree of skewness of the data distribution. Positive skewness means that the data deviates to the right from the mean, and negative skewness means that the data deviates to the left from the mean. The shape of the data distribution remains basically unchanged over time, so the skewness will not change significantly when the data is skewed to the left or right; kurtosis describes the degree of peaks in the data distribution, that is, the measurement of the concentration of the data. The kurtosis of data in a stable state will not change significantly either, because the concentration and distribution shape of the data remain basically unchanged over time, so it is meaningless to extract skewness and kurtosis features.

[0086] In an embodiment of the present invention, the multidimensional data features of the current curve in each of the interval segments are aggregated into a data feature set of the current curve, including: taking the maximum value, waveform factor and kurtosis calculated according to the current curve in the first stage and the third stage as the first data feature; taking the average value, standard deviation, root mean square, maximum value, extreme value, waveform factor, pulse factor and mode calculated according to the current curve in the second stage as the second data feature; and aggregating the first data feature and the second data feature into the data feature set of the current curve.

[0087] In detail, the average value is calculated by the following formula:

[0088]

[0089] Wherein, p(b) represents the bth current data included in the current curve, and f1 represents the average value of a current data included in the current curve.

[0090] The standard deviation is calculated using the following formula:

[0091]

[0092] Wherein, p(b) represents the bth current data included in the current curve, f1 represents the average value of a current data included in the current curve, and f2 represents the standard deviation of the current curve.

[0093] The root mean square is calculated by the following formula:

[0094]

[0095] Among them, p 2 (b) represents the square of the bth current data included in the current curve, and f3 represents the root mean square of a current data included in the current curve.

[0096] The form factor is calculated using the following formula:

[0097]

[0098] Among them, p 2 (b) represents the square of the bth current data included in the current curve,

[0099] |p(b)| represents the absolute value of the bth current data included in the current curve, and f6 represents the waveform factor of the ath current data included in the current curve.

[0100] The pulse factor is calculated by the following formula:

[0101]

[0102] Among them, max(p(b)) represents the maximum current value in the current curve, |p(b)| represents the absolute value of the bth current data contained in the current curve, and f7 represents the pulse factor of a current data contained in the current curve.

[0103] The skewness is calculated using the following formula:

[0104]

[0105] Among them, p(b) represents the bth current data included in the current curve, f1 represents the average value of a current data included in the current curve, and f8 represents the skewness of a current data included in the current curve.

[0106] The kurtosis is calculated by the following formula:

[0107]

[0108] Wherein, p(b) represents the bth current data contained in the current curve, f1 represents the average value of a current data contained in the current curve, and f9 represents the kurtosis of a current data contained in the current curve.

[0109] The mode is the current data with the largest number of occurrences among the a current data contained in the current curve, the maximum is the current data with the largest number among the a current data contained in the current curve, and the maximum and minimum difference is calculated by subtracting the minimum value from the maximum value among the a current data contained in the current curve.

[0110] Aggregating the multidimensional data features of the current curve in each interval segment into a data feature set is conducive to in-depth analysis and comparison of the behaviors of different interval segments, and is conducive to identifying and understanding the changing trends, statistical properties and spectral information of the current characteristics in each interval segment, thereby providing a more targeted data basis for anomaly detection, pattern recognition or predictive modeling.

[0111] S5. Perform clustering calculation on the data feature set using a preset objective function, and determine the state of the turnout where the target turnout device is located according to the calculation result.

[0112] In an embodiment of the present invention, the clustering calculation of the data feature set using a preset objective function includes: converting the data feature set into a numerical matrix; initializing the numerical matrix using random numbers; calculating the cluster center of each data in the data feature set; updating the initialized numerical matrix using the cluster center to obtain an updated matrix; iteratively calculating the updated matrix using a preset objective function to obtain a membership matrix; and determining the clustering result of each data point in the data feature set based on the membership matrix.

[0113] In detail, the data feature set is clustered by a fuzzy clustering algorithm. The fuzzy clustering algorithm is an algorithm that uses membership to determine the degree of clustering to which each data point belongs. It helps to optimize the clustering process, improve the efficiency and accuracy of clustering, and ensure that the generated clusters can better judge the state of the turnout where the target turnout equipment is located. By using a preset objective function, a more appropriate clustering method can be selected to obtain a clustering result that is more conducive to analysis.

[0114] In an embodiment of the present invention, determining the state of the turnout where the target turnout device is located based on the calculation results includes: selecting target data one by one from the data feature set, and obtaining reference data points in the cluster corresponding to the target data; calculating a first average distance between the target data and the reference data points; determining adjacent clusters of the cluster corresponding to the target data, calculating the average distance between the target data and all data points in the adjacent clusters, and obtaining a second average distance; calculating the contour coefficient of the target data in the membership matrix one by one according to the first average distance and the second average distance; calculating the average contour coefficient according to the contour coefficient of the target data; if the average contour coefficient is greater than a preset contour coefficient, determining that the state of the turnout where the target turnout device is located is a healthy state; if the average contour coefficient is equal to the preset contour coefficient, determining that the state of the turnout where the target turnout device is located is a sub-healthy state; if the average contour coefficient is less than the preset contour coefficient, determining that the state of the turnout where the target turnout device is located is a faulty state.

[0115] In detail, the average silhouette coefficient is an indicator used to evaluate the quality of clustering. It combines the average distance between a data feature set and the cluster to which the set is assigned and the average distance between the set and other clusters. It is used to measure the compactness and separation of clusters. The state of the turnout where the target turnout equipment is located can be judged by the density and separation of the clusters, providing a more effective method for the diagnosis of turnout equipment.

[0116] The data feature set X={x1,x 2,...., x n The fuzzy set of} can be represented by the membership matrix U = [u ij ] means that the element u of matrix U ij Indicates the membership degree of the jth (j=1.2...c) cluster, u ij The following conditions must be met: This condition means that the element u ij The sum of the membership degrees belonging to all clustering degrees is 1.

[0117] The preset objective function is:

[0118]

[0119] d ij (x j ,v i )=||v i -x j ||

[0120]

[0121] Among them, m represents the weighted index, U represents the membership matrix, V represents the cluster center, n represents the number of data in the data feature set, c represents the membership of the cth clustering degree, u ij It represents the degree of membership of the jth (j=1,2,...,c) cluster, d ij is the distance from the data in the current curve corresponding to the data feature set to the cluster center, x j represents the jth data in the data feature set, v i represents the i-th cluster center, (min)J m (U,V) represents the minimum value of the objective function.

[0122] The constraints that the objective function needs to satisfy are:

[0123]

[0124] Among them, n represents the number of data in the data feature set, c represents the membership degree of the cth clustering degree, and uij It represents the degree of membership to the jth (j=1,2,...,c) cluster.

[0125] In detail, the cluster centers and memberships are iteratively calculated by a preset objective function until the objective function values ​​of the N+1th iteration and the Nth iteration differ by 0.001, or the number of iterations is exceeded, and the calculation stops. When the objective function is calculated to a minimum value, when the objective function changes less than the number of iterations (inflection point), the optimal number of clusters is determined, and convergence is judged at this time to obtain a membership matrix.

[0126] The objective function represents the sum of the products of the membership of all the data feature sets and the distances from the data feature sets to each cluster center.

[0127] Through cluster analysis and membership matrix, target turnout equipment in similar conditions can be classified into the same cluster, thereby identifying the general working status or potential problems of the equipment, timely discovering abnormal conditions of the equipment, predicting possible failures in advance, and facilitating the adoption of preventive maintenance measures, improving equipment reliability and availability, reducing maintenance costs and downtime, and optimizing the operational management efficiency of turnout equipment.

[0128] like Figure 2 As shown, it is a functional module diagram of a fault diagnosis system based on turnout current characteristics provided by an embodiment of the present invention. According to the functions implemented, the fault diagnosis system based on turnout current characteristics 400 may include a data acquisition module 401, a data cleaning module 402, a region division module 403, a feature extraction module 404 and a data analysis module 405. The module described in the present invention may also be referred to as a unit, which refers to a series of computer program segments that can be executed by a processor of the fault diagnosis system based on turnout current characteristics and can complete fixed functions, which are stored in the memory of the fault diagnosis system based on turnout current characteristics.

[0129] In this embodiment, the functions of each module / unit are as follows:

[0130] A data acquisition module, used to acquire an action current data set of a target turnout device within a preset time;

[0131] A data cleaning module, used for classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and generating a turnout current data set according to the cleaned data and the normal data;

[0132] A region division module, used for generating a current curve according to the turnout current data set, and dividing the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve;

[0133] A feature extraction module, used for aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve;

[0134] The data analysis module is used to perform clustering calculation on the data feature set using a preset objective function, and determine the state of the turnout where the target turnout equipment is located according to the calculation result.

[0135] In detail, the modules described in the turnout current characteristic fault diagnosis system 400 described in the embodiment of the present invention adopt the same technical means as the turnout current characteristic fault diagnosis method described in the accompanying drawings when used, and can produce the same technical effects, which will not be repeated here.

[0136] like Figure 3 FIG. 1 is a schematic diagram of the structure of an electronic device for implementing a fault diagnosis method based on turnout current characteristics provided by an embodiment of the present invention.

[0137] The electronic device 1000 may include a processor 1001, a memory 1002, a communication bus 1003 and a communication interface 1004, and may also include a computer program stored in the memory 1002 and executable on the processor 1001, such as a program for generating a fault diagnosis model based on turnout current characteristics.

[0138] Among them, the processor 1001 is the control core (Control Unit) of the electronic device, which uses various interfaces and lines to connect the various components of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing programs or modules stored in the memory 1002, and calling data stored in the memory 1002.

[0139] The memory 1002 includes at least one type of readable storage medium. In some embodiments, the memory 1002 may be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory 1002 may be used to store not only application software installed in the electronic device but also various data.

[0140] The communication bus 1003 is configured to implement connection and communication between the memory 1002 and at least one processor 1001, etc.

[0141] The communication interface 1004 is used for communication between the electronic device and other devices, including a network interface and a user interface.

[0142] The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device 1000, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0143] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0144] The program for generating a system model for fault diagnosis based on turnout current characteristics stored in the memory 1002 of the electronic device 1000 is a combination of multiple instructions. When running in the processor 1001, the steps in the above-mentioned method for fault diagnosis based on turnout current characteristics can be implemented.

[0145] Specifically, the specific implementation system of the processor 1001 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0146] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the steps in the above-mentioned method for diagnosing faults based on turnout current characteristics can be implemented.

[0147] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0148] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0149] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A fault diagnosis method based on turnout current characteristics, characterized in that: The method comprises: Obtaining an action current data set of a target turnout device within a preset time; Classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and generating a turnout current data set based on the cleaned data and the normal data; generating a current curve according to the turnout current data set, and dividing the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve; Aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve; The data feature set is clustered using a preset objective function, and the state of the turnout where the target turnout device is located is determined based on the calculation result.

2. A fault diagnosis method based on turnout current characteristics according to claim 1, characterized in that: The classifying the action current data set into abnormal data and normal data includes: Calculating a standard deviation of the action current data set; The data in the action current data set that is larger than the standard deviation is regarded as abnormal data; The data in the action current data set that is less than or equal to the standard deviation is regarded as normal data.

3. A fault diagnosis method based on turnout current characteristics according to claim 2, characterized in that: The performing data cleaning on the abnormal data includes: Filling the null value data in the abnormal data by using an interpolation algorithm; The non-null data in the abnormal data are replaced by the current mean value in the normal data.

4. A fault diagnosis method based on turnout current characteristics according to claim 1, characterized in that: The step of dividing the current curve into a plurality of preset interval segments according to the first-order derivative of the current curve comprises: Obtaining a state conversion timestamp of the turnout where the turnout device is located, and removing corresponding data points on the current curve according to the state conversion timestamp to obtain an optimized curve; Calculating the first-order derivative of each point on the optimization curve; Selecting, in chronological order, a first data point and a second data point on the optimization curve whose first-order derivative is less than a preset derivative threshold; dividing the curve before the first data point on the current curve into the first stage of the current curve; dividing a curve between the first data point and the second data point on the current curve into a second stage of the current curve; The portion of the current curve after the second data point is divided into a third stage of the current curve.

5. A fault diagnosis method based on turnout current characteristics according to claim 4, characterized in that: The multi-dimensional data features of the current curve in each of the interval segments are aggregated into a data feature set of the current curve, including: taking the maximum value, the waveform factor and the kurtosis calculated according to the current curve in the first stage and the third stage as the first data feature; The average value, standard deviation, root mean square, maximum value, extreme value, waveform factor, pulse factor and mode calculated according to the current curve in the second stage are used as second data features; The first data feature and the second data feature are aggregated into a data feature set of the current curve.

6. A fault diagnosis method based on turnout current characteristics according to claim 5, characterized in that: The performing clustering calculation on the data feature set by using a preset objective function includes: Converting the data feature set into a numerical matrix; Initializing the numerical matrix using random numbers; Calculate the cluster center of each data in the data feature set; Using the cluster centers to update the initialized numerical matrix to obtain an updated matrix; Iteratively calculating the update matrix using a preset objective function to obtain a membership matrix; The clustering result of each data point in the data feature set is determined according to the membership matrix.

7. A fault diagnosis method based on turnout current characteristics according to claim 6, characterized in that: The step of determining the state of the turnout where the target turnout device is located according to the calculation result includes: Selecting target data one by one from the data feature set, and obtaining reference data points in the cluster corresponding to the target data; Calculating a first average distance between the target data and the reference data points; Determine the adjacent clusters of the cluster corresponding to the target data, calculate the average distance between the target data and all data points in the adjacent clusters, and obtain a second average distance; Calculate the silhouette coefficient of the target data in the membership matrix one by one according to the first average distance and the second average distance; Calculate an average silhouette coefficient based on the silhouette coefficient of the target data; If the average contour coefficient is greater than a preset contour coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a healthy state; If the average profile coefficient is equal to the preset profile coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a sub-healthy state; If the average contour coefficient is less than a preset contour coefficient, it is determined that the state of the turnout where the target turnout equipment is located is a fault state.

8. A fault diagnosis system based on turnout current characteristics, characterized in that: The system comprises: A data acquisition module, used to acquire an action current data set of a target turnout device within a preset time; A data cleaning module, used for classifying the action current data set into abnormal data and normal data, performing data cleaning on the abnormal data, and generating a turnout current data set according to the cleaned data and the normal data; A region division module, used for generating a current curve according to the turnout current data set, and dividing the current curve into a plurality of preset interval segments according to a first-order derivative of the current curve; A feature extraction module, used for aggregating the multi-dimensional data features of the current curve in each of the interval segments into a data feature set of the current curve; The data analysis module is used to perform clustering calculation on the data feature set using a preset objective function, and determine the state of the turnout where the target turnout equipment is located according to the calculation result.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute a fault diagnosis method based on turnout current characteristics as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, a fault diagnosis method based on turnout current characteristics as claimed in any one of claims 1 to 7 is implemented.