Compensation capacitor state automatic identification method and device

Abstract

The application discloses a kind of compensation capacitance state automatic identification method and device, the method includes: obtaining signal equipment detection data, the signal equipment detection data includes track circuit detection data and compensation capacitance detection data;The track circuit detection data is sectioned, obtains multiple small envelope signals of each section;The compensation capacitance detection data is divided according to track circuit section, obtains the compensation capacitance detection data of each section;The multiple small envelope signals of each section are extracted and selected, and the track circuit feature data of each section is obtained, which represents the state of compensation capacitance;The compensation capacitance detection data of each section is identified, and the pulse recognition result of compensation capacitance detection data is obtained;The state of compensation capacitance is identified by combining the pulse recognition result of compensation capacitance detection data and the track circuit feature data.The application can realize the state of compensation capacitance identification with high efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of dynamic detection of railway equipment and fault diagnosis of equipment, and in particular to a method and device for automatically identifying the state of compensation capacitors. BACKGROUND

[0002] This section is intended to provide background or context to the embodiments of the application recited in the claims. The description herein does not constitute admission that the prior publication, square, or subject matter described herein and / or the material contained therein is or was prior art to the claims at issue.

[0003] With the continuous development of high-speed railways, higher requirements are put forward for the maintenance and repair of existing signal equipment. As a crucial part of the train operation control system, the track circuit is mainly responsible for monitoring the occupation, position and integrity of the train. During signal transmission, the track circuit produces reactive power loss due to the inductance of the steel rail, which leads to insufficient effective transmission distance and aggravates the influence of the "skin" effect of the signal during transmission on the steel rail. Therefore, it is necessary to install compensation capacitors at equal intervals in the track circuit section to compensate for the inductance of the steel rail in sections, maximize the adverse effects of the inductance of the steel rail on transmission, and make the characteristic impedance of the track circuit tend to be resistive, so as to ensure the improvement of the "signal-to-noise ratio" and the transmission length of the track circuit, the inspection of the broken rail of the steel rail, and the increase of the shunt current of the locomotive signal at the entrance end of the track circuit. When the compensation capacitor fails, the receiving voltage of the track circuit receiver will decrease, and when multiple compensation capacitors in the same section fail simultaneously, the corresponding track relay in the section may fall down, causing a "red light band" failure and affecting the transportation efficiency.

[0004] Currently, the compensation capacitor state of high-speed and conventional speed railways is periodically detected by using comprehensive detection vehicles and electric power detection vehicles. The commonly used method is to analyze the detection data in real time by relying on manual experience to confirm the operating condition of the compensation capacitors along the line. However, with the rapid growth of the railway transportation network, the detection tasks of the comprehensive detection vehicles and electric power detection vehicles of the whole railway are increasing, and the amount of detection data is also expanding. Relying on manual fault troubleshooting by detection personnel often consumes a lot of manpower and resources, and it is difficult to meet the needs of the field.

[0005] Therefore, there is a need for an efficient automatic identification scheme for the state of compensation capacitors. SUMMARY

[0006] The embodiments of the present application provide an automatic identification method for the state of compensation capacitors, which realizes efficient automatic identification of the state of compensation capacitors. The method comprises:

[0007] Obtaining signal equipment detection data, the signal equipment detection data comprising track circuit detection data and compensation capacitor detection data;

[0008] The track circuit detection data is divided into sections to obtain a plurality of small envelope signals of each section;

[0009] The compensation capacitor detection data is divided into sections according to the track circuit to obtain compensation capacitor detection data of each section;

[0010] The plurality of small envelope signals of each section are subjected to feature extraction and selection to obtain track circuit feature data of each section representing the compensation capacitor state;

[0011] The compensation capacitor detection data of each section is subjected to pulse recognition of the compensation capacitor detection data to obtain a pulse recognition result of the compensation capacitor detection data;

[0012] The compensation capacitor state is recognized in combination with the pulse recognition result of the compensation capacitor detection data and the track circuit feature data.

[0013] The embodiment of the application provides a compensation capacitor state automatic recognition device to realize efficient compensation capacitor state automatic recognition, which comprises:

[0014] A detection data obtaining module is configured to obtain signal equipment detection data, wherein the signal equipment detection data comprises track circuit detection data and compensation capacitor detection data;

[0015] A section dividing module is configured to divide the track circuit detection data into sections to obtain a plurality of small envelope signals of each section, and divide the compensation capacitor detection data into sections according to the track circuit to obtain compensation capacitor detection data of each section;

[0016] A feature extraction and selection module is configured to extract and select the plurality of small envelope signals of each section to obtain track circuit feature data of each section representing the compensation capacitor state;

[0017] A pulse recognition module is configured to recognize the compensation capacitor detection data of each section to obtain a pulse recognition result of the compensation capacitor detection data;

[0018] A compensation capacitor state recognition module is configured to recognize the compensation capacitor state in combination with the pulse recognition result of the compensation capacitor detection data and the track circuit feature data.

[0019] The embodiment of the application further provides a computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor implements the above-mentioned compensation capacitor state automatic recognition method when executing the computer program.

[0020] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the automatic compensation capacitor state identification method.

[0021] The embodiment of the present application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to realize the automatic compensation capacitor state identification method.

[0022] In the embodiment of the present application, signal equipment detection data is obtained, the signal equipment detection data comprises track circuit detection data and compensation capacitor detection data; the track circuit detection data is divided into sections, and a plurality of small envelope signals of each section are obtained; the compensation capacitor detection data is divided according to track circuit sections, and compensation capacitor detection data of each section is obtained; a plurality of small envelope signals of each section are subjected to feature extraction and selection, and track circuit feature data of each section representing a compensation capacitor state is obtained; the compensation capacitor detection data of each section is subjected to pulse identification of compensation capacitor detection data, and a pulse identification result of the compensation capacitor detection data is obtained; the pulse identification result of the compensation capacitor detection data and the track circuit feature data are combined to identify a compensation capacitor state. The method provided in the embodiment of the present application combines the track circuit detection data and the compensation capacitor detection data, extracts the track circuit feature data of each section representing the compensation capacitor state, obtains the pulse identification result of the compensation capacitor detection data, and then performs fusion analysis, so as to comprehensively judge the compensation capacitor state along the railway, realize automatic identification of the state, and improve the work efficiency and reduce the work pressure of personnel. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor. In the drawings:

[0024] Figure 1 The flow chart of the automatic compensation capacitor state identification method in the embodiment of the present application;

[0025] Figure 2 The structural schematic diagram of the railway signal dynamic detection system in the embodiment of the present application;

[0026] Figure 3 The schematic diagram of the signal equipment detection data in the embodiment of the present application;

[0027] Figure 4A schematic diagram of automatic recognition of insulating joints by using intersection method in the embodiment of the present application;

[0028] Figure 5 A layout rule of compensation capacitors in the embodiment of the present application;

[0029] Figure 6 A schematic diagram of automatic recognition of compensation capacitor states in the embodiment of the present application; Figure 5 A corresponding short-circuit voltage curve;

[0030] Figure 7 A specific flow chart of pulse recognition of compensation capacitor detection data in the embodiment of the present application;

[0031] Figure 8 A specific flow chart of compensation capacitor state recognition in the embodiment of the present application;

[0032] Figure 9 A schematic diagram of automatic recognition device of compensation capacitor states in the embodiment of the present application;

[0033] Figure 10 A schematic diagram of a computer device in the embodiment of the present application. DETAILED DESCRIPTION

[0034] To make the purpose, technical scheme and advantages of the embodiment of the present application more clear, the embodiment of the present application is further described in detail below with reference to the drawings. Herein, the schematic embodiment of the present application and its description are used to explain the present application, but not as a limitation to the present application.

[0035] Figure 1 A flow chart of automatic recognition method of compensation capacitor states in the embodiment of the present application, as shown in Figure 1 , the method comprises:

[0036] Step 101, obtaining signal device detection data, the signal device detection data comprising track circuit detection data and compensation capacitor detection data;

[0037] Step 102, dividing the track circuit detection data into sections to obtain a plurality of small envelope signals of each section;

[0038] Step 103, dividing the compensation capacitor detection data according to track circuit sections to obtain compensation capacitor detection data of each section;

[0039] Step 104, performing feature extraction and selection on a plurality of small envelope signals of each section to obtain track circuit feature data of each section representing compensation capacitor states;

[0040] Step 105, performing pulse recognition of compensation capacitor detection data on compensation capacitor detection data of each section to obtain pulse recognition results of compensation capacitor detection data;

[0041] In step 106, the pulse recognition result of the compensation capacitor detection data is combined with the track circuit feature data to perform compensation capacitor state recognition.

[0042] In the embodiment of the present application, the track circuit feature data of each section representing the compensation capacitor state is extracted in combination with the track circuit detection data and the compensation capacitor detection data, the pulse recognition result of the compensation capacitor detection data is obtained, and then fusion analysis is performed to comprehensively judge the compensation capacitor state along the railway, so that automatic recognition of the compensation capacitor state is realized, and the work efficiency on site is improved and the work pressure of personnel is reduced.

[0043] In step 101, signal equipment detection data is obtained, and the signal equipment detection data includes track circuit detection data and compensation capacitor detection data; wherein the signal equipment detection data can be collected by a railway signal dynamic detection system; the signal equipment detection data includes not only the track circuit detection data and the compensation capacitor detection data, but also traction return current detection data and balise detection data; the railway signal dynamic detection system refers to signal detection equipment based on a railway high-speed comprehensive detection vehicle and a signal detection vehicle platform, which is composed of track circuit, compensation capacitor, traction return current, balise and other detection subsystems to realize all-around detection of track circuit transmission characteristics, frequency characteristics, track circuit compensation capacitor, point balise and locomotive signal operation state under the condition of train dynamic operation. Figure 2 FIG. 1 is a structural schematic diagram of the railway signal dynamic detection system in the embodiment of the present application. The signal equipment detection data refers to railway signal track circuit (short-circuit voltage, carrier frequency, low frequency), compensation capacitor, traction return current, balise detection data obtained based on the signal dynamic detection system, which is displayed in the same mileage coordinate, Figure 3 FIG. 2 is a schematic diagram of the signal equipment detection data in the embodiment of the present application. The track circuit in the embodiment of the present application refers to a circuit composed of a steel rail line and a steel rail insulation along the railway, which is used to automatically and continuously detect whether the line is occupied by a locomotive vehicle and to send train control information to the train. The track circuit takes a section as a basic unit, and adjacent sections are separated by mechanical insulation joints or electrical insulation joints, and the carrier frequency of the same section is consistent.

[0044] In step 102, the track circuit detection data is divided into sections to obtain a plurality of small envelope signals of each section; that is, based on the track circuit detection data shown in FIG. 3, the track circuit detection data of each section is automatically divided based on an insulation joint automatic recognition method, so that the track circuit detection data of each section is obtained. Figure 3

[0045] In an embodiment, the track circuit detection data is divided to obtain a plurality of small envelope signals of each section, including:

[0046] ​The track circuit detection data is divided by track circuit section based on the insulating section automatic identification method, and track circuit detection data of each section is obtained;

[0047] The track circuit detection data of each section is normalized.

[0048] The normalized track circuit detection data of each section is divided again according to the number of section capacitances in combination with the signal basic library information, and a plurality of small envelope signals of each section are obtained.

[0049] In an embodiment, the method further comprises:

[0050] The insulating section is automatically identified by using the intersection method, and the identified insulating section is obtained.

[0051] Figure 4 It is a schematic diagram of the intersection method for automatically identifying the insulating section in the embodiment of the application. f1 is the carrier frequency voltage curve of the previous track circuit section, and f2 is the carrier frequency voltage curve of the next track circuit section. When the voltage of f2 is greater than f1, and the voltage of f1 is less than 0.1v at this time, it is considered that the point is the starting point of the next section, that is, D point is the starting point of f2 section, and C point is the end point of f1 section, so as to realize the automatic division of track circuit section.

[0052] In the above embodiment, the track circuit detection data of each section is normalized to ensure that the track circuit detection data result of each section is between [0, 1]. The calculation formula is as follows:

[0053]

[0054] Among them, represents the minimum value of the estimated value in time t, represents the maximum value of in time t.

[0055] In step 103, the compensation capacitance detection data is divided according to the track circuit section, and the compensation capacitance detection data of each section is obtained. The method is similar to the method of step 102. The compensation capacitance detection data is divided according to the track circuit section based on the insulating section automatic identification method, and the compensation capacitance detection data of each section is obtained.

[0056] In the above embodiments, based on the signal base library information, the normalized track circuit detection data for each segment is further divided according to the number of capacitors in the segment, resulting in multiple small envelope signals for each segment. The signal base library information refers to the basic library information required by the railway signal dynamic detection system, mainly including information such as station, signal, segment carrier frequency, low frequency, number of compensation capacitors, kilometer markers, and segment length. The number of compensation capacitors indicates the actual number of compensation capacitors that should be present in the track circuit segment. In actual lines, compensation capacitors are arranged at equal intervals. Figure 5 This invention provides the arrangement rules for the compensation capacitors in this embodiment. Zr represents the equivalent impedance of the rail, C represents the compensation capacitor, and lc represents the distance between adjacent capacitors. The distance between the capacitors at both ends of the section and the insulating joint is lc / 2, while the distance between the remaining capacitors is lc. Under the condition of a short circuit in the track circuit, Figure 6 In the embodiments of the present invention, and Figure 5 The corresponding short-circuit voltage curves, and the multiple envelope curves obtained after secondary division according to the actual location of the compensation capacitor, are shown below. Figure 6 The curve between the two dashed lines.

[0057] In step 104, feature extraction and selection are performed on multiple small envelope signals of each segment to obtain track circuit feature data of each segment representing the state of the compensation capacitor.

[0058] In one embodiment, feature extraction and selection are performed on multiple small envelope signals of each segment to obtain track circuit feature data characterizing the state of the compensation capacitor for each segment, including:

[0059] Feature extraction is performed on multiple small envelope signals in each segment to obtain the extracted features;

[0060] The extracted features are used to construct a feature vector space, and a feature selection method based on the Fisher criterion function is used to obtain track circuit feature data representing the state of the compensation capacitor for each segment.

[0061] In one embodiment, the extracted features include one or any combination of time-domain statistical parameters such as inlet / outlet difference, maximum / minimum difference, maximum value, minimum value, average value, kurtosis, fluctuation factor, peak factor, difference sum, etc.

[0062] In the above embodiments, the track circuit feature data refers to the multidimensional feature data of each envelope obtained after feature extraction and selection, which is the track circuit feature that can effectively characterize the state of the compensation capacitor.

[0063] In step 105, pulse recognition is performed on the compensation capacitor detection data of each segment to obtain the pulse recognition result of the compensation capacitor detection data.

[0064] In an embodiment, before the pulse recognition of the compensation capacitance detection data of each section, further comprising:

[0065] The compensation capacitance detection data of each section is normalized and denoised to obtain processed compensation capacitance detection data;

[0066] The pulse recognition of the compensation capacitance detection data of each section comprises:

[0067] The pulse recognition of the processed compensation capacitance detection data of each section is performed.

[0068] In the above embodiment, the normalization and denoising refer to normalizing the compensation capacitance detection data to ensure that the compensation capacitance detection data of each section is between [0, 1], and the calculation method refers to formula (1). Then, the denoising is performed, and the median filter denoising method is used to denoise the section compensation capacitance to process the interference of factors such as step signals and sudden signals on the recognition result. The calculation formula is as follows:

[0069] y(i)'=median(y(i-N),y(i-N+1),…,y(i),y(i+1),…,y(i+N)) (2)

[0070] Where y(i)' represents the result of median filtering of the ith value, 2N+1 is the width of the median filter window function, and y(i-N), y(i-N+1), …, y(i), y(i+1), …, y(i+N) are signal samples in the window, where y(i) is the signal sample value at the center of the window.

[0071] Figure 7 The specific flowchart for the pulse recognition of the compensation capacitance detection data in the embodiment of the application is shown in Figure 7 The principle is to automatically recognize the pulse by the adaptive threshold method. In an embodiment, the pulse recognition of the compensation capacitance detection data of each section comprises:

[0072] For each section, the compensation capacitance detection data of the section is divided into multiple intervals and interval lengths according to the compensation capacitance number N of the section; wherein the compensation capacitance detection data is represented by a set Data={data1,data2,……,datan}, wherein the number of sections is n; after equal interval division, the length of each interval is w;

[0073] The pulse recognition threshold Thr of the section is calculated;

[0074] According to the pulse recognition threshold of the section, calculate the compensation capacitance detection data mean value datamean(k) of each interval in the section;

[0075] For each interval in the section, determine whether each point datai in the interval exceeds the pulse recognition threshold Thr of the section or not, where datamean(k) is the compensation capacitance detection data mean value of the interval;

[0076] If yes, record the points exceeding the pulse recognition threshold and the corresponding position index; where the record can be represented as TData={(Tdata1,Indexdata1),(Tdata2,Indexdata2),…,(Tdatam,Indexdatam)}, m is the interval number;

[0077] Divide the corresponding position index of each point in the record by the interval length w of the interval, and perform rounding processing to obtain the pulse recognition result of the compensation capacitance detection data.

[0078] In the above embodiment, when equal-interval division is performed, let n be the number of section data points, and N be the number of compensation capacitances of the section, then the length of the first N-1 sections is equal-interval length w=floor(n / N), where floor is the floor function, and the number of points of the last section is n-(N-1)×w.

[0079] In an embodiment, the pulse recognition threshold Thr of the section is calculated, including:

[0080] For each interval k of the section, calculate the compensation capacitance detection data mean value datamean(k) and the compensation capacitance detection data maximum value datamax(k) of the interval, respectively;

[0081] Subtract the compensation capacitance detection data mean value from the compensation capacitance detection data maximum value to obtain the threshold Thr of the interval k k , and add the threshold set THR={Thr1, Thr2, …, Thr N};

[0082] Take the median of the threshold set THR as the pulse recognition threshold of the section, that is, Thr=meadian(THR).

[0083] In step 106, combine the pulse recognition result of the compensation capacitance detection data and the track circuit characteristic data to perform compensation capacitance state recognition.

[0084] Figure 8 For the specific flowchart of the compensation capacitance state recognition in the embodiment of the present application, see Figure 8In an embodiment, the pulse recognition result of the compensation capacitor detection data and the track circuit feature data are combined to perform compensation capacitor state recognition, including:

[0085] Based on the pulse recognition result of the compensation capacitor detection data, it is determined whether the pulse signal at the capacitor position is normally recognized;

[0086] If yes, it is determined that the compensation capacitor state is normal;

[0087] If no, it is determined whether the track circuit feature data exceeds the track circuit typical feature threshold value;

[0088] If yes, it is determined that the compensation capacitor state is failure;

[0089] If no, it is determined that the compensation capacitor state is normal.

[0090] The track circuit typical feature threshold value judgment is based on a large amount of historical data, and is obtained by analyzing the distribution of the track circuit typical features in normal and fault states. In an embodiment, the method further includes:

[0091] The track circuit typical feature threshold value is determined by the following steps:

[0092] Obtain typical line track circuit detection data and compensation capacitor fault operation and maintenance data;

[0093] Divide the typical line track circuit detection data into sections to obtain a track circuit section typical feature matrix; wherein the track circuit section typical feature matrix can be represented as follows:

[0094]

[0095] wherein represents the pth feature of the qth interval, N is the number of compensation capacitors of the ith section, and k is the dimension of the track circuit feature data;

[0096] According to the compensation capacitor fault operation and maintenance data, the track circuit section typical feature matrix is classified to obtain normal sample data and fault sample data;

[0097] The k-means clustering algorithm is used to cluster the normal sample data and the fault sample data respectively to obtain the clustering centers of the normal sample data and the fault sample data respectively;

[0098] The Euclidean distance of each sample data in the normal sample data and the fault sample data to the corresponding clustering center is calculated respectively, and is sorted from small to large to obtain a first distance set of the normal sample data to the corresponding clustering center a second distance set of the fault sample data to the corresponding clustering center

[0099] Selecting the first set The preset quantile value in the middle is taken as a typical feature threshold value of the track circuit;Wherein m, n respectively represent the number of normal samples and fault samples.

[0100] Wherein, the preset quantile value can be 95%, and the 95% quantile value in L1 is taken as a judgment threshold value, that is, the point whose distance from the center of the normal sample exceeds the 95% quantile value of the sample data is considered as the compensation capacitor failure.

[0101]

[0102] In summary, the method proposed in the embodiment of the application has the following advantages:

[0103] Firstly, the state of the compensation capacitor in the real-time detection process can be automatically identified, the algorithm calculation speed and identification accuracy are high, the efficiency of the field staff can be effectively improved, and the work pressure can be reduced.

[0104] Secondly, the insulating joint automatic identification method based on the intersection method can effectively and quickly realize the automatic division of the track circuit section detection data.

[0105] Thirdly, the compensation capacitor pulse automatic identification method based on the pulse identification threshold value can quickly identify the pulse signal state at the compensation capacitor position by using the compensation capacitor detection data.

[0106] Fourthly, the track circuit typical feature threshold value calculation method based on the K-means clustering algorithm can automatically calculate the clustering center of the normal fault data and the fault sample data by using the typical line track circuit detection data and the compensation capacitor fault operation and maintenance data, so that the calculation of the track circuit typical feature threshold value can be quickly realized. Meanwhile, the method has strong applicability, and can be used for the calculation of different track circuit typical feature threshold values of different lines, so as to meet the requirement of large difference of detection data of different lines.

[0107] The embodiment of the application also proposes a compensation capacitor state automatic identification device, which has similar principles to the compensation capacitor state automatic identification method, and will not be described here.

[0108] Figure 9 Fig. 1 is a schematic diagram of a compensation capacitor state automatic identification device in the embodiment of the application, which comprises:

[0109] The detection data obtaining module 901 is used for obtaining signal equipment detection data, and the signal equipment detection data comprises track circuit detection data and compensation capacitor detection data.

[0110] The section division module 902 is configured to divide the track circuit detection data by sections to obtain a plurality of small envelope signals of each section; and divide the compensation capacitance detection data by track circuit sections to obtain compensation capacitance detection data of each section.

[0111] The feature extraction and selection module 903 is configured to extract and select features from the plurality of small envelope signals of each section to obtain track circuit feature data of each section representing a compensation capacitance state.

[0112] The pulse recognition module 904 is configured to recognize pulses in the compensation capacitance detection data of each section to obtain a pulse recognition result of the compensation capacitance detection data.

[0113] The compensation capacitance state recognition module 905 is configured to recognize a compensation capacitance state in combination with the pulse recognition result of the compensation capacitance detection data and the track circuit feature data.

[0114] In an embodiment, the section division module 902 is specifically configured to:

[0115] Divide the track circuit detection data by track circuit sections based on an automatic insulating joint recognition method to obtain track circuit detection data of each section.

[0116] Perform normalization processing on the track circuit detection data of each section.

[0117] In combination with signal base library information, perform secondary division on the normalized track circuit detection data of each section according to the number of section capacitances to obtain a plurality of small envelope signals of each section.

[0118] In an embodiment, the section division module 902 is further configured to:

[0119] Perform automatic insulating joint recognition using an intersection method to obtain recognized insulating joints.

[0120] In an embodiment, the feature extraction and selection module 903 is specifically configured to:

[0121] Extract features from the plurality of small envelope signals of each section to obtain extracted features.

[0122] Construct a feature vector space using the extracted features, and perform feature selection using a method based on a Fisher criterion function to obtain track circuit feature data of each section representing a compensation capacitance state.

[0123] In an embodiment, the extracted features include one or any combination of an inlet-outlet difference, a maximum-minimum difference, a maximum value, a minimum value, an average value, a kurtosis, a fluctuation factor, a peak factor, a difference, and an isochronous domain statistical parameter.

[0124] In an embodiment, the section division module 902 is specifically configured to:

[0125] Before the pulse identification of the compensation capacitance detection data of each section, the compensation capacitance detection data of each section is normalized and denoised to obtain processed compensation capacitance detection data;

[0126] The pulse identification module 904 is specifically configured to:

[0127] The pulse identification of the compensation capacitance detection data is performed on the processed compensation capacitance detection data of each section.

[0128] In an embodiment, the pulse identification module 904 is specifically configured to:

[0129] For each section, the compensation capacitance detection data of the section is divided into multiple intervals and interval lengths according to the compensation capacitance number of the section;

[0130] The pulse identification threshold of the section is calculated;

[0131] The compensation capacitance detection data mean of each interval in the section is calculated according to the pulse identification threshold of the section;

[0132] For each interval in the section, it is determined whether each point in the interval exceeds the compensation capacitance detection data mean of the interval and the pulse identification threshold of the section;

[0133] If yes, the points exceeding the pulse identification threshold and the corresponding position indexes are recorded;

[0134] The corresponding position index of each point is divided by the interval length of the interval, and rounding processing is performed to obtain a value as the pulse identification result of the compensation capacitance detection data.

[0135] In an embodiment, the pulse identification module 904 is specifically configured to:

[0136] For each interval of the section, the compensation capacitance detection data mean and the compensation capacitance detection data maximum of the interval are calculated respectively;

[0137] The compensation capacitance detection data mean and the compensation capacitance detection data maximum are subtracted to obtain a threshold value of the interval, and the threshold value is added to a threshold value set;

[0138] The median of the threshold value set is taken as the pulse identification threshold of the section.

[0139] In an embodiment, the compensation capacitance state identification module 905 is specifically configured to:

[0140] Based on the pulse recognition result of the compensation capacitor detection data, it is judged whether the pulse signal at the capacitor position is normally recognized or not.

[0141] If yes, it is determined that the compensation capacitor state is normal.

[0142] If no, it is judged whether the track circuit feature data exceeds the track circuit typical feature threshold value or not.

[0143] If yes, it is determined that the compensation capacitor state is failure.

[0144] If no, it is determined that the compensation capacitor state is normal.

[0145] In an embodiment, the compensation capacitor state recognition module 905 is further used for:

[0146] The track circuit typical feature threshold value is determined by the following steps:

[0147] Typical line track circuit detection data and compensation capacitor fault operation and maintenance data are obtained.

[0148] The typical line track circuit detection data is divided into sections to obtain a track circuit section typical feature matrix.

[0149] According to the compensation capacitor fault operation and maintenance data, the track circuit section typical feature matrix is classified to obtain normal sample data and fault sample data.

[0150] The k-means clustering algorithm is used to cluster the normal sample data and the fault sample data respectively to obtain the clustering centers of the normal sample data and the fault sample data respectively.

[0151] The Euclidean distance of each sample data in the normal sample data and the fault sample data to the corresponding clustering center is calculated respectively, and is sorted from small to large to obtain a first distance set of the normal sample data to the corresponding clustering center and a second distance set of the fault sample data to the corresponding clustering center.

[0152] The preset quantile value in the first set is selected as the track circuit typical feature threshold value.

[0153] In summary, the device proposed in the embodiment of the application has the following beneficial effects:

[0154] Firstly, the automatic discrimination of the compensation capacitor state in the real-time detection process can be realized, and the algorithm calculation speed and recognition accuracy are high, which can effectively improve the efficiency of the field staff and reduce the work pressure.

[0155] Secondly, the insulating joint automatic recognition method based on the intersection method can effectively and quickly realize the automatic division of the track circuit section detection data.

[0156] Third, the compensation capacitor pulse automatic identification method based on the pulse identification threshold can quickly identify the pulse signal state at the compensation capacitor position by using the compensation capacitor detection data.

[0157] Fourth, the track circuit typical feature threshold calculation method based on the K-means clustering algorithm can automatically calculate the clustering center of normal fault data and fault sample data by using typical line track circuit detection data and compensation capacitor fault operation and maintenance data, and can quickly calculate the track circuit typical feature threshold. Meanwhile, the method has strong applicability, and can calculate different track circuit typical feature thresholds for different lines to meet the requirement of large difference of detection data of different lines.

[0158] The embodiment of the present application also provides a computer device, Figure 10 The computer device 1000 includes a memory 1010, a processor 1020 and a computer program 1030 stored in the memory 1010 and executable on the processor 1020, and the processor 1020 implements the above compensation capacitor state automatic identification method when executing the computer program 1030.

[0159] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above compensation capacitor state automatic identification method.

[0160] The embodiment of the present application also provides a computer program product, which includes a computer program, and the computer program is executed by a processor to implement the above compensation capacitor state automatic identification method.

[0161] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can adopt a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0162] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0163] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0164] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0165] The above-described specific embodiments are merely intended to further describe the purpose, technical solutions and beneficial effects of the present application, and it should be understood that the above-described specific embodiments are merely specific embodiments of the present application and are not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for compensating for automatic identification of a state of a capacitor, characterized by, The method comprises the following steps: obtaining signal device detection data, wherein the signal device detection data comprises track circuit detection data and compensation capacitor detection data; segmenting the track circuit detection data to obtain a plurality of small envelope signals of each segment; dividing the compensation capacitor detection data according to track circuit segments to obtain compensation capacitor detection data of each segment; extracting and selecting a plurality of small envelope signals of each segment to obtain track circuit feature data of each segment representing the compensation capacitor state; performing pulse recognition on the compensation capacitor detection data of each segment to obtain a pulse recognition result of the compensation capacitor detection data; combining the pulse recognition result of the compensation capacitor detection data and the track circuit feature data to perform compensation capacitor state recognition; performing pulse recognition on the compensation capacitor detection data of each segment, comprising: for each segment, dividing the compensation capacitor detection data of the segment at equal intervals according to the number of compensation capacitors of the segment to obtain a plurality of intervals and interval lengths; calculating a pulse recognition threshold of the segment; calculating the average value of the compensation capacitor detection data of each interval in the segment according to the pulse recognition threshold of the segment; for each interval in the segment, determining whether each point in the interval exceeds the average value of the compensation capacitor detection data of the interval; if yes, recording the point and the corresponding position index; dividing the corresponding position index of each recorded point by the interval length of the interval and performing rounding processing to obtain a value as the pulse recognition result of the compensation capacitor detection data; combining the pulse recognition result of the compensation capacitor detection data and the track circuit feature data to perform compensation capacitor state recognition, comprising: based on the pulse recognition result of the compensation capacitor detection data, determining whether the pulse signal at the capacitor position is normally recognized; if yes, determining that the compensation capacitor state is normal; if no, determining whether the track circuit feature data exceeds a typical track circuit feature threshold; if yes, determining that the compensation capacitor state is invalid; if no, determining that the compensation capacitor state is normal.

2. The method of claim 1, wherein, dividing the track circuit detection data to obtain a plurality of small envelope signals of each segment, comprising: dividing the track circuit detection data according to track circuit segments based on an automatic insulating joint recognition method to obtain track circuit detection data of each segment; performing normalization processing on the track circuit detection data of each segment; combining signal base library information to perform secondary division on the normalized track circuit detection data of each segment according to the number of segment capacitors to obtain a plurality of small envelope signals of each segment.

3. The method of claim 1, wherein, Further comprising: using the intersection method to automatically recognize insulating joints to obtain recognized insulating joints.

4. The method of claim 1, wherein, extracting and selecting a plurality of small envelope signals of each segment to obtain track circuit feature data of each segment representing the compensation capacitor state, comprising: extracting features from a plurality of small envelope signals of each segment to obtain extracted features; The feature vector space is constructed using the extracted features, a feature selection method based on a Fisher criterion function is used, and track circuit feature data of each section representing the state of the compensation capacitor is obtained.

5. The method of claim 4, wherein, The extracted features include one or any combination of an inlet-outlet difference, a maximum-minimum difference, a maximum value, a minimum value, an average value, a kurtosis, a fluctuation factor, a peak factor, a difference, and an isochronous domain statistical parameter.

6. The method of claim 1, wherein, Before the compensation capacitor detection data of each section is subjected to pulse identification, the method further includes: normalizing and denoising the compensation capacitor detection data of each section to obtain processed compensation capacitor detection data; Before the compensation capacitor detection data of each section is subjected to pulse identification, the method further includes: normalizing and denoising the compensation capacitor detection data of each section to obtain processed compensation capacitor detection data; 7. The method of claim 1, wherein, Before the compensation capacitor detection data of each section is subjected to pulse identification, the method further includes: normalizing and denoising the compensation capacitor detection data of each section to obtain processed compensation capacitor detection data; The pulse identification threshold of the section is calculated, including: the mean value and the maximum value of the compensation capacitor detection data of each interval of the section are calculated respectively; 8. The method of claim 1, wherein, the mean value and the maximum value of the compensation capacitor detection data are subtracted to obtain the threshold value of the interval, and the threshold value is added to the threshold value set; the median of the threshold value set is taken as the pulse identification threshold of the section. The method further includes: The track circuit typical feature threshold is determined by the following steps: obtaining typical line track circuit detection data and compensation capacitor fault operation and maintenance data; dividing the typical line track circuit detection data into sections to obtain a track circuit section typical feature matrix; classifying the track circuit section typical feature matrix according to the compensation capacitor fault operation and maintenance data to obtain normal sample data and fault sample data; using a k-means clustering algorithm to cluster the normal sample data and the fault sample data respectively to obtain clustering centers of the normal sample data and the fault sample data respectively; 9. A compensation capacitance state automatic recognition device characterized by comprising: calculating the Euclidean distance of each sample data in the normal sample data and the fault sample data to the corresponding clustering center respectively, and sorting from small to large to obtain a first distance set of the normal sample data to the corresponding clustering center and a second distance set of the fault sample data to the corresponding clustering center; selecting a preset quantile value in the first set as the track circuit typical feature threshold. The method includes: a detection data obtaining module configured to obtain signal equipment detection data, the signal equipment detection data including track circuit detection data and compensation capacitor detection data; a section division module configured to divide the track circuit detection data into sections to obtain a plurality of small envelope signals of each section; the compensation capacitor detection data is divided according to track circuit sections to obtain compensation capacitor detection data of each section; a feature extraction and selection module configured to extract and select features from the plurality of small envelope signals of each section to obtain track circuit feature data of each section representing the state of the compensation capacitor; a pulse identification module configured to identify pulses in the compensation capacitor detection data of each section to obtain pulse identification results of the compensation capacitor detection data. The compensation capacitor state recognition module is configured to recognize the compensation capacitor state in combination with the pulse recognition result of the compensation capacitor detection data and the track circuit characteristic data. The pulse recognition module is specifically configured to: divide the compensation capacitor detection data of each section at equal intervals according to the number of compensation capacitors of the section, to obtain a plurality of intervals and interval lengths; calculate a pulse recognition threshold of the section; calculate a compensation capacitor detection data mean of each interval in the section according to the pulse recognition threshold of the section; determine whether each point in each interval in the section exceeds the pulse recognition threshold of the section; if yes, record the point and a corresponding position index; divide the corresponding position index of each recorded point by the interval length of the interval, and perform rounding processing to obtain a pulse recognition result of the compensation capacitor detection data; The compensation capacitor state recognition module is specifically configured to: determine whether the pulse signal at the capacitor position is normally recognized based on the pulse recognition result of the compensation capacitor detection data; if yes, determine that the compensation capacitor state is normal; if no, determine whether the track circuit characteristic data exceeds the track circuit typical characteristic threshold; if yes, determine that the compensation capacitor state is invalid; if no, determine that the compensation capacitor state is normal.

10. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the method in any one of claims 1 to 8.

11. A computer readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 8.

12. A computer program product, characterised in that, The computer program product includes a computer program, and the computer program is executed by the processor to implement the method in any one of claims 1 to 8.

Images