An interval track circuit electrical characteristic analysis and fault diagnosis method, device, electronic equipment and computer readable medium
By leveraging big data and machine learning technologies, the electrical characteristics of track circuits are analyzed in real time. Combined with manual diagnosis, the knowledge base is automatically updated, solving the problems of insufficient scalability and timeliness in existing technologies, and enabling efficient diagnosis and maintenance of track circuit faults.
Patent Information
- Application Number
- CN202111047378.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-09-08
- Publication Date
- 2026-01-20
- Estimated Expiration
- 2041-09-08
AI Technical Summary
Existing track circuit monitoring and fault analysis methods suffer from poor scalability and timeliness, making it difficult to effectively address the complex fault modes of track circuits and the impact of environmental factors.
By employing big data analytics and machine learning methods, and through the collaborative work of a host computer and a server, the system collects electrical characteristic data of track circuits in real time, performs anomaly detection and feature matching, and combines this with manual diagnosis to automatically update the track circuit knowledge base, thereby achieving autonomous learning and fault analysis.
It improves the scalability and timeliness of track circuit fault analysis, enabling timely detection of safety hazards and providing scientific and reliable maintenance and repair support.
Smart Images
Figure CN115795116B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of rail transit monitoring and analysis, in particular to an interval track circuit electrical characteristic analysis and fault diagnosis method and device, electronic equipment and computer readable medium. BACKGROUND
[0002] Track circuit is one of the three major parts of outdoor railway signal equipment, which bears the important responsibility of ensuring the safe operation of railways, and is the basis of railway train operation control system. Track circuits in China have experienced a development process from direct current pulse track circuit to alternating current continuous track circuit. With the rapid development of high-speed railways, ZPW2000 series track circuits are widely used in signal systems of high-speed railways, passenger dedicated lines and existing line reconstruction in China. At present, the mainstream model of interval track circuit is ZPW2000-A and ZPW2000-R series, which accounts for more than 99%.
[0003] The maintenance of track circuit has always been the key task of railway signal maintenance department, and its fault proportion accounts for more than 30% of the faults of railway signal system. Especially for high-speed railways, the safety pressure is great, and the safe and stable operation of ZPW2000 series track circuit is particularly important. In recent years, with the development of monitoring technology, the technical department of the railway general company has installed various sensors indoors and outdoors of track circuit for collecting operation parameters of track circuit. How to make good use of these data and find out the real fault cause from these data is the key research direction in recent years.
[0004] In related technologies, the monitoring and fault analysis of track circuit are mainly based on the analysis method of fault tree model. These methods cannot be separated from signal professionals, and usually track circuit experts edit fault case processes according to various forms of track circuit faults, and hand over them to software developers for implementation. The advantage is that the data conforming to the fault characteristics can be accurately analyzed, but it is proved in practice that this method has the characteristics of poor expansibility and difficult case collection. When a new fault case occurs, the fault process and software need to be edited, and the on-site diagnosis software needs to be upgraded, which has poor timeliness. The operation of track circuit is greatly affected by environmental factors such as weather and temperature, as well as factors such as cable, line and track bed. It is difficult to include various fault forms based on the fault tree analysis model.
[0005] Therefore, the applicant believes that the existing track circuit monitoring and fault analysis method has the defects of poor expansibility and poor timeliness, and needs to be improved. SUMMARY
[0006] In order to improve the problem of poor expansibility and poor timeliness of the track circuit monitoring and fault analysis method, the present application provides an interval track circuit electrical characteristic analysis and fault diagnosis method, device, electronic equipment and computer readable medium.
[0007] In a first aspect, the application provides a method for analyzing electrical characteristics and diagnosing faults of an interval track circuit, which adopts the following technical solution:
[0008] A method for analyzing electrical characteristics and diagnosing faults of an interval track circuit is applied to an upper computer, and includes the following steps:
[0009] A download request is sent to a server according to a preset condition, wherein the download request is used to request to download a track circuit knowledge base;
[0010] Receipt data of the server is received, and the track circuit knowledge base at the upper computer end is updated according to the receipt data;
[0011] Real-time electrical characteristic data of track circuit operation sent by a lower computer is acquired;
[0012] It is judged whether the acquired real-time electrical characteristic data is abnormal;
[0013] If there is an abnormality, an abnormal data feature is extracted, matched with data features in the track circuit knowledge base at the upper computer end, and a knowledge base abnormal case set conforming to an abnormal working state feature is obtained;
[0014] The knowledge base abnormal case set is analyzed and first information is obtained, and the abnormal data feature and the first information are sent to the server, so that the server performs track circuit fault analysis training and updates the track circuit knowledge base at the server end according to the abnormal data feature and the first information, wherein the first information includes a first fault reason and a matching degree obtained by the upper computer analysis, and the track circuit fault analysis training is performed when the matching degree of the first fault reason is greater than a set threshold;
[0015] The abnormal data feature and second information obtained after artificial diagnosis are sent to the server, so that the server performs track circuit fault analysis training and updates the track circuit knowledge base at the server end according to the abnormal data feature and the second information, wherein the second information includes a second fault reason obtained by artificial diagnosis.
[0016] In some embodiments, the data feature is a data feature of a single parameter of the track circuit or a combined data feature of multiple parameters, and is represented as: t1, t2…t n The data feature includes a feature weight and a feature ratio, the feature weight refers to a weight of the parameter in judging the fault reason, and the feature ratio refers to a change amplitude of the parameter relative to a standard value.
[0017] In some embodiments, the track circuit knowledge base includes multiple sets of feature case data sets, represented as: M={ M1, M2…M m}, wherein Mi ={t 1i ,t 2i …t ni}, i∈[1,m], where m is the number of feature case datasets; each set M i This is a characteristic case of a track circuit, including data characteristics and a description of the cause of the fault; the abnormal data characteristics are represented as: P={t 1p ,t 2p …t np}
[0018] In some implementations, the method for obtaining the first fault cause and matching degree includes the following steps:
[0019] The parameter t in the abnormal data feature P is compared with the corresponding parameter t in the feature case data set M to obtain a comparison result set Q. The comparison result set Q includes the feature weight information of parameter t, whether parameter t matches, and the offset matching value.
[0020] Based on feature weights and offset matching values, several sets are selected from the comparison result set Q to obtain the knowledge base anomaly case set Q1~Q2. n ;
[0021] For the knowledge base anomaly case set Q1~Q n According to the expression: k = ((∑(t)) 权重 &t 匹配 )) / t 个数 () * 100%), to get Q1~Q n The corresponding matching degree k is sorted according to the matching degree to obtain several first fault causes and matching degrees.
[0022] In some implementations, the track circuit fault analysis training includes the following steps:
[0023] Based on the first fault cause and abnormal data feature P, a feature case M is matched to the track circuit knowledge base on the server side. i The abnormal data feature P is included in the feature case M according to the weighted average ratio skew algorithm. i In the process, it completes self-learning training and automatic updates of the track circuit knowledge base on the server side;
[0024] Based on the second fault cause and abnormal data feature P, if the second fault cause is already included in the track circuit knowledge base on the server side, then the feature case M in the track circuit knowledge base on the server side will be matched. i If the second cause of failure is not included in the track circuit knowledge base on the server side, a new feature case M is formed. i; the abnormal data feature P is counted into the feature case M according to a weighted average ratio inclination algorithm i In the method, autonomous learning training is completed and automatic updating of the track circuit knowledge base on the server side is completed.
[0025] In some embodiments, the weighted average ratio inclination algorithm comprises the following steps:
[0026] t 1p in the abnormal data feature P is selected i t 1i in the feature case M 1i权重 is obtained according to an expression: t 1p = (t 1i权重 * sample number) / (sample number + 1) i t 1i new weight data in the feature case M
[0027] A new sample number is obtained according to an expression: sample number = sample number + 1
[0028] t 2p to t np in the abnormal data feature P are sequentially selected, t i to t 2i in the feature case M ni correspond, and new weight data of each feature value in the feature case M i is obtained according to the above expression.
[0029] In a second aspect, the application provides an interval track circuit electrical characteristic analysis and fault diagnosis method, which adopts the following technical scheme:
[0030] An interval track circuit electrical characteristic analysis and fault diagnosis method is applied to a server and comprises the following steps:
[0031] A download request of an upper computer is received, wherein the download request is used to request to download a track circuit knowledge base.
[0032] Acknowledgement data is sent to the upper computer, so that the upper computer updates the track circuit knowledge base on the upper computer side according to the acknowledgement data.
[0033] Abnormal data features and first information sent by the upper computer are received, wherein the first information comprises first fault reasons and matching degrees obtained by the upper computer through analysis on a knowledge base abnormal case set, the knowledge base abnormal case set is obtained by the upper computer through extraction of abnormal data features and matching of the abnormal data features with feature data in the track circuit knowledge base on the upper computer side, and the abnormal data features are obtained by the upper computer through acquisition of real-time electrical characteristic data of track circuit work sent by the lower computer and judgment on whether there is an abnormality.
[0034] If the matching degree of the first fault cause is greater than a set threshold, track circuit fault analysis training is performed, and the track circuit knowledge base on the server side is updated; otherwise, manual diagnosis is entered;
[0035] The abnormal data features and second information sent by the upper computer are received, wherein the second information includes a second fault cause obtained through manual diagnosis;
[0036] If the second fault cause is already included in the track circuit knowledge base on the server side, track circuit fault analysis training is performed, and the track circuit knowledge base on the server side is updated; if the second fault cause is not included in the track circuit knowledge base on the server side, track circuit fault analysis training is performed, a new feature case is formed, and the track circuit knowledge base on the server side is updated.
[0037] In some embodiments, the server includes several servers, and the several servers perform track circuit knowledge base merging according to feature data and version information in the track circuit knowledge base on the server side.
[0038] In a third aspect, the application provides an interval track circuit electrical characteristic analysis and fault diagnosis device, which adopts the following technical scheme:
[0039] An interval track circuit electrical characteristic analysis and fault diagnosis device includes
[0040] A download request sending module is configured to send a download request to a server according to a preset condition, wherein the download request is used to request downloading a track circuit knowledge base.
[0041] A receipt data receiving module is configured to receive receipt data of the server and update the track circuit knowledge base on the upper computer side according to the receipt data.
[0042] A real-time data obtaining module is configured to obtain electrical characteristic real-time data of track circuit work sent by a lower computer.
[0043] An abnormality judging module is configured to judge whether the obtained electrical characteristic real-time data is abnormal.
[0044] A feature data matching module is configured to extract abnormal data features when there is abnormality, match the abnormal data features with data features in the track circuit knowledge base on the upper computer side, and obtain a knowledge base abnormal case set conforming to abnormal working state features.
[0045] The first information sending module is configured to analyze the knowledge base abnormal case set and obtain first information, and send the abnormal data features and the first information to the server, so that the server performs track circuit fault analysis training according to the abnormal data features and the first information and updates the track circuit knowledge base on the server side, wherein the first information includes a first fault cause and a matching degree obtained by the upper computer analysis, and the track circuit fault analysis training is performed when the matching degree of the first fault cause is greater than a set threshold value.
[0046] The second information sending module is configured to send the abnormal data features and second information obtained after artificial diagnosis to the server, so that the server performs track circuit fault analysis training according to the abnormal data features and the second information and updates the track circuit knowledge base on the server side, wherein the second information includes a second fault cause obtained by artificial diagnosis.
[0047] In a fourth aspect, the application provides an interval track circuit electrical characteristic analysis and fault diagnosis device, which adopts the following technical scheme:
[0048] An interval track circuit electrical characteristic analysis and fault diagnosis device, which comprises
[0049] The download request receiving module is configured to receive a download request of the upper computer, wherein the download request is used to request to download the track circuit knowledge base.
[0050] The receipt data sending module is configured to send receipt data to the upper computer, so that the upper computer updates the track circuit knowledge base on the upper computer side according to the receipt data.
[0051] The first information receiving module is configured to receive the abnormal data features and the first information sent by the upper computer, wherein the first information includes a first fault cause and a matching degree obtained by the upper computer analyzing the knowledge base abnormal case set, the knowledge base abnormal case set is obtained by the upper computer extracting abnormal data features and matching the feature data in the track circuit knowledge base on the upper computer side, and the abnormal data features are obtained by the upper computer acquiring real-time electrical characteristic data of track circuit operation sent by the lower computer and judging whether there is an abnormality.
[0052] The first information judging module is configured to judge whether the matching degree of the first fault cause in the first information is greater than a set threshold value, if yes, perform track circuit fault analysis training, and update the track circuit knowledge base on the server side, otherwise, turn to artificial diagnosis.
[0053] The second information receiving module is configured to receive the abnormal data features and the second information sent by the upper computer, wherein the second information includes a second fault cause obtained by artificial diagnosis.
[0054] The second information judging module is configured to judge whether the second fault cause is contained in the track circuit knowledge base on the server side, if yes, perform track circuit fault analysis training and update the track circuit knowledge base on the server side, and if not, perform track circuit fault analysis training, form a new feature case, and update the track circuit knowledge base on the server side.
[0055] In a fifth aspect, the present application provides an electronic device, which adopts the technical scheme as follows:
[0056] An electronic device comprises
[0057] at least one processor;
[0058] a storage device configured to store at least one computer program;
[0059] When the at least one computer program is executed by the at least one processor, the at least one processor implements the above-mentioned interval track circuit electrical characteristic analysis and fault diagnosis method.
[0060] In a sixth aspect, the present application provides a computer readable medium, which adopts the technical scheme as follows:
[0061] A computer readable medium has a computer program stored thereon, and the computer program is configured to implement the above-mentioned interval track circuit electrical characteristic analysis and fault diagnosis method when executed by a processor.
[0062] In summary, the present application has at least one of the following beneficial technical effects:
[0063] 1. The method discards the traditional signal professional knowledge, applies the big data analysis and machine learning method, analyzes the track circuit operation electrical characteristic data from the mathematical and probabilistic point of view, and obtains an interval track circuit electrical characteristic analysis and fault diagnosis method which is self-learning from zero.
[0064] 2. Through the electrical characteristic analysis and fault diagnosis of the section track circuit, the safety hidden danger of the interval track circuit in the operation process is found in time, and the fault cause is analyzed and judged when the track circuit fails, so as to provide scientific and reliable technical support for the maintenance and repair of the track circuit, and the method has strong expansibility and timeliness. BRIEF DESCRIPTION OF DRAWINGS
[0065] Figure 1 A basic data flow diagram of the host computer and the server provided by the embodiment of the present application is provided;
[0066] Figure 2 A basic working flow diagram of the host computer provided by the embodiment of the present application is provided;
[0067] Figure 3The matching flowchart of the abnormal data features of the host computer and the case set of the knowledge base is provided for the embodiment of the present application.
[0068] Figure 4 The basic working flowchart of the server is provided for the embodiment of the present application.
[0069] Figure 5 The flowchart of the server receiving the new case to update the track circuit knowledge base is provided for the embodiment of the present application.
[0070] Figure 6 The flowchart of the track circuit knowledge base information of two servers being merged to form a new track circuit knowledge base is provided for the embodiment of the present application. DETAILED DESCRIPTION
[0071] The following will be described in detail with reference to the accompanying drawings. Figures 1 to 6 The present application will be further described in detail.
[0072] The embodiment of the present application discloses a method for analyzing electrical characteristics and diagnosing faults of an interval track circuit. The method comprises a lower computer, a host computer and a server. The lower computer is used for collecting real-time data of electrical characteristics of the track circuit. The host computer is an analysis station computer. The server is a track circuit knowledge base server.
[0073] As shown in the basic data flowchart of the analysis station computer and the track circuit knowledge base server in the embodiment, the figure describes the timing relationship of the command and data interaction between the substation computer and the server. Figure 1 As shown in the basic working flowchart of the substation computer in the embodiment, the following steps are included at the analysis station computer end.
[0074] As shown in the basic working flowchart of the substation computer in the embodiment, the following steps are included at the analysis station computer end. Figure 2
[0075] A1: periodically downloading the knowledge database from the track circuit knowledge base server;
[0076] A1.1: the analysis station computer applies for the track circuit knowledge base to the server according to certain rules, including program starting, server connection recovery, timing (the high-speed rail window repair time is considered in the embodiment, and the setting is 1:30 in the evening) and the like. The application information contains the version number of the knowledge database at the analysis station computer end.
[0077] A1.2: the analysis station computer receives the reply data of the server, judges the data type of the server, and if the type is the version number, it is considered that the server is consistent with the current analysis station computer version, and there is no need to update. If the type is the data, that is, the knowledge base content data, the knowledge base at the analysis station computer end is updated.
[0078] A1.3: Analyzing the configuration items for logical judgment from the knowledge base: mainly the weight proportion setting item, the electrical characteristic parameter offset proportion setting item, the output reason number configuration item, and the normal working electrical characteristic value of the track circuit. The configured parameters in this implementation case are: weight proportion = 70%, electrical characteristic parameter offset proportion = 5%, and output reason number = 3.
[0079] A2: Obtaining the real-time working electrical characteristic data of the track circuit from the data input interface. The data input interface is an open protocol that can be transmitted by different track circuit interface devices and can support multiple interface modes such as serial port, network, CAN port, etc. In this implementation case, the network TCP protocol is used to obtain data from the track circuit maintenance host. The obtained data is organized in the format of "type, serial number, value" and provided to the data analysis module.
[0080] A3: Judging whether the real-time data of the current track circuit is normal data or abnormal data based on the signal logic relationship and data offset proportion;
[0081] A3.1: According to the three-point logical check relationship of the automatic block train control of the railway signal, the track circuit is divided into two working states: normal train operation and idle operation. The real-time data is classified according to these two working states.
[0082] A3.2: Reading the user-set data deviation proportion of the device and the standard electrical characteristic value of the track circuit of the device from the configuration file;
[0083] A3.3: According to the normal train operation and idle operation working states, respectively calculating whether the real-time data of each parameter is within the data offset proportion range of the standard electrical characteristic value of the track circuit (the deviation range is set to 5% in this implementation example). When one item is outside the offset range, it is considered that the current track circuit is working abnormally and is stored in the abnormal state buffer.
[0084] A3.4: Looping A3.1~A3.3 until all parameter data analysis is complete.
[0085] A3.5: If the abnormal state buffer is not empty, it is considered that the track circuit has an abnormal state and enters the A4 working process. Otherwise, exit the loop and proceed to the data judgment of the next track section.
[0086] A4: As shown in Figure 3 , analyzing the data of the abnormal working state, extracting the offset standard value data parameter, and matching it with the characteristic data of the current knowledge base. A set of fault reason collection that meets the characteristics of the abnormal working state is obtained from the knowledge base. The specific steps are as follows:
[0087] A4.1: Extract the deviation parameter t in the abnormal working state data, compare it with the working state value of parameter t in the current knowledge base, find the parameters consistent with the deviation (the deviation ratio of this embodiment is set to 5%), and obtain a data knowledge case M that meets the deviation ratio i ;
[0088] A4.2: Repeat the step of A4.1, when 70% of the deviation ratio of parameter t is consistent with the case M i , store the knowledge base case M i in the buffer zone that meets the buffer zone;
[0089] A4.3: Repeat the steps of A4.1~A4.2 to obtain a set of comparison result set Q that meets the current abnormal characteristic data.
[0090] A5: Analyze the knowledge base case set obtained in the step of A4, comprehensively consider the compliance degree and weight of the parameters, output the fault reason, and upload to the track circuit knowledge base server, the specific steps are as follows:
[0091] A5.1: Extract the data characteristics of the abnormal data according to the distribution of the data, numbered as: t1, t2…t n ;
[0092] A5.2: Further, match the data characteristics extracted from the abnormal data with the comparison result set Q, and pick out the best n sets from Q, called knowledge base abnormal case set Q1~Q n ;
[0093] A5.3: Further, pick out the three M n with the highest matching ratio from Q1~Q i , calculate the probability of the fault reason matching degree, and output the fault reason, the output track circuit abnormal reason information form is: according to the number of data samples n, the possibility of reason M1 is k1%, the possibility of reason M2 is k2%, and the possibility of reason M3 is k n %;
[0094] A5.4: Further, upload the fault characteristic data and the output reason information to the track circuit knowledge base server.
[0095] A6: The system provides a man-machine interface (I / O interface of the upper computer), the user determines or modifies the fault reason through the man-machine interface, and uploads the fault data and the diagnosis result to the server after the user intervention;
[0096] A6.1: When the highest probability of the abnormal or fault reason output in the step of A5 is greater than 70%, the user can actively modify the fault reason information through the man-machine interface, and after the modification is completed, the modified reason information is automatically uploaded to the server;
[0097] A6.2: Further, when the probability of the output abnormal or fault cause is less than or equal to 70%, the system actively pops up a human - machine interface dialog box, requests the user to confirm or modify the fault cause, and after completion, automatically uploads the modified cause information to the server.
[0098] Figure 4 The following shows the basic working flow chart of the track circuit knowledge base server in this embodiment. Figure 5 This is the schematic diagram of the process for the server in this embodiment to receive new feature cases and update the track circuit knowledge base. The following steps are included at the track circuit knowledge base server side:
[0099] B: The server receives the request from the analysis station machine to download the track circuit knowledge base, and judges whether the version number of the request is the same as the currently used version. If it is the same, it returns the current version number; otherwise, it returns the track circuit knowledge base. The track circuit knowledge base includes content such as a fault case knowledge base, parameter setting information, and normal working standard values of the track circuit.
[0100] B2: The server receives the abnormal working electrical characteristic data transmitted by the analysis station machine. When the fault cause of the received track circuit is a case included in the knowledge base and the ratio exceeds 70%, it starts the track circuit fault analysis training algorithm, calculates the proportion of various parameter data of this type of fault, and automatically updates the track circuit knowledge base, including the following steps:
[0101] B2.1: Further, the track circuit fault analysis training algorithm, according to the analysis result and data feature P = {t 1p , t 2p …t np} of this time, matches and finds M i in the track circuit knowledge base. According to {t i , t 1i , t 2i …t ni} in the knowledge base M i , the data feature P of this time is incorporated into the knowledge base M
[0102] B2.2: According to the expression: t 1i权重 = (t 1p + t 1i权重 * number of samples) / (number of samples + 1), obtain the new weight data of t i in M 1i ;
[0103] B2.3: According to the expression: number of samples = number of samples + 1, obtain the new number of samples;
[0104] B2.4: Sequentially select t in P2p t np , respectively corresponding to the feature case M i t 2i t ni , referring to the expressions in B2.2 and B2.3 above, the new weight data of each feature value in the feature case M i is obtained. After all the feature parameters are calculated, the weight information of each parameter t in the track circuit knowledge base M i is updated, and the automatic learning, training and automatic updating of the knowledge base of the machine are completed.
[0105] B3: The server receives the abnormal working electrical characteristic data transmitted by the analysis station, and when the fault cause of the received track circuit is not included in the current track circuit knowledge base, the track circuit fault analysis training algorithm is started, the data weight of this type of fault is automatically analyzed, and a new fault model (feature case) is formed and stored in the current track circuit knowledge base, realizing the self-learning and accumulation process of the track circuit knowledge base from zero, and the specific steps are as follows:
[0106] B3.1: A new fault model M i , first analyze the similar parameters of a single parameter in M i , form a new feature parameter t, t can be a single parameter or a combination of multiple parameters;
[0107] B3.2: Further, the similar parameters refer to a group of parameters that have the same fluctuation amplitude or sudden drop ratio compared with the normal working state value;
[0108] B3.3: Further, a new fault model, the fault data weight is compared with the track circuit standard value according to the feature parameter t, the greater the deviation, the greater the weight, indicating that the characteristics of the data are more obvious;
[0109] B3.4: Train a data model from zero, the more the number of samples accumulated in the training analysis, the more accurate the analysis result can be obtained.
[0110] B4: Each different server can adopt online or offline mode to merge the track circuit knowledge base, integrate the track circuit knowledge base in each server according to the version information and the description of the knowledge base feature data, and update the weight information of each parameter in the track circuit knowledge base and the training sample number information, as shown in Figure 6 , this embodiment describes the present application by merging two different track circuit knowledge bases, including the following steps:
[0111] B4.1: Extract the basic version number of the server track circuit knowledge base, and obtain the sample data in each server track circuit knowledge base, select one version with more sample data and newer time as the basic version, in this case, No. 1 knowledge base is taken as the basic version;
[0112] B4.2: Extract a case M in No. 2 knowledge base i Compare with the feature case in the basic version, match the time, place, data parameter ratio and other information contained in the sample data, if it is the same sample data, do not repeat the analysis and calculation;
[0113] B4.3: Further, if it is the same type of case, but contains different basic information, then the case M i Integrate the information of the case with the information of the case in the basic library, recalculate the weight of the parameter t and other information, and update the related information of the case in the basic version;
[0114] B4.4: Further, if it is a new case type, then add the case type to the basic version library;
[0115] B4.5: Repeat steps B4.2~B4.4 until all cases in No. 2 knowledge base have been successfully merged into the basic version library.
[0116] B5: When the track circuit knowledge base is changed, update the version number of the track circuit knowledge base, and record the log information of the version update.
[0117] In some possible implementation manners, the electronic device according to the embodiments of the present application can at least include at least one processor and at least one storage device. The storage device stores at least one computer program, and when the computer program is executed by the processor, the processor executes the steps in the method according to various specific embodiments of the present application described in the above technical solutions in the specification.
[0118] In some possible implementation manners, various aspects of the present application can also be implemented as a computer readable medium having a computer program stored thereon, which is used to implement the steps in the method according to various specific embodiments of the present application described in the above technical solutions in the specification when the computer program is executed by the processor of the electronic device.
[0119] It should be noted that the above-mentioned medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.
[0120] The readable signal medium can include a data signal propagated in a baseband or as a part of a carrier wave, in which readable program codes are borne. Such a propagated data signal can take on various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The readable signal medium can also be any readable medium other than the readable storage medium, which can send, propagate or transmit programs for use by or in connection with an instruction execution system, apparatus or device.
[0121] The program code contained on the readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination of the above.
[0122] The program code for performing the operations of the present application can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, etc., and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's electronic device, partly on the user's electronic device and partly on a remote electronic device, or entirely on the remote electronic device or server. In situations in which the remote electronic device is involved, the remote electronic device can be connected to the user's electronic device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external electronic device (for example, through an Internet service provider to connect to the Internet).
[0123] Those skilled in the art can understand that various aspects of the present application can be implemented as a system, a method or a program product. Therefore, various aspects of the present application can be embodied as a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, which can be collectively referred to as "circuitry", "module" or "system" here.
[0124] The above are all preferred embodiments of the present application, and do not limit the protection scope of the present application, so that: all equivalent changes made according to the structure, shape, principle of the present application should be covered in the protection scope of the present application.
Claims
1. A method for analyzing the electrical characteristics and diagnosing faults in a track circuit, applied in a host computer, characterized in that, Includes the following steps: A download request is sent to the server according to preset conditions, wherein the download request is used to request the download of the track circuit knowledge base; Receive the receipt data from the server and update the track circuit knowledge base on the host computer based on the receipt data; Acquire real-time electrical characteristic data of the track circuit operation sent by the lower-level machine; Determine if there are any anomalies in the acquired real-time electrical characteristic data; If an anomaly is found, the abnormal data features are extracted and matched with the data features in the track circuit knowledge base of the host computer to obtain a set of abnormal cases in the knowledge base that conform to the abnormal working state features. The abnormal case set of the knowledge base is analyzed to obtain first information. The abnormal data features and the first information are sent to the server so that the server can perform track circuit fault analysis training based on the abnormal data features and the first information and update the track circuit knowledge base on the server side. The first information includes the first fault cause and matching degree obtained by the host computer analysis. The track circuit fault analysis training is carried out when the matching degree of the first fault cause is greater than a set threshold. The abnormal data features and the second information obtained after manual diagnosis are sent to the server so that the server can perform track circuit fault analysis training based on the abnormal data features and the second information and update the track circuit knowledge base on the server side. The second information includes the second fault cause obtained by manual diagnosis.
2. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 1, characterized in that: The data features are either single-parameter data features of the track circuit or combined data features of multiple parameters, represented as: t1, t2…t n The data features include feature weights and feature ratios. The feature weights refer to the weight a parameter carries when determining the cause of a fault, and the feature ratios refer to the magnitude of a parameter’s change relative to a standard value.
3. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 2, characterized in that: The track circuit knowledge base includes multiple sets of feature case data, represented as: M={ M1,M2…M m }, where M i ={t 1i ,t 2i …t ni }, i∈[1,m], where m is the number of feature case datasets; each set M i This is a characteristic case of a track circuit, including data characteristics and a description of the cause of the fault; the abnormal data characteristics are represented as: P={t 1p ,t 2p …t np } 4. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 3, characterized in that: The method for obtaining the first fault cause and matching degree includes the following steps: The parameter t in the abnormal data feature P is compared with the corresponding parameter t in the feature case data set M to obtain a comparison result set Q. The comparison result set Q includes the feature weight information of parameter t, whether parameter t matches, and the offset matching value. Based on feature weights and offset matching values, several sets are selected from the comparison result set Q to obtain the knowledge base anomaly case set Q1~Q2. n ; For the knowledge base anomaly case set Q1~Q n According to the expression: k = ((∑(t)) 权重 &t 匹配 )) / t 个数 ) * 100%, get Q1~Q n The corresponding matching degree k is sorted according to the matching degree to obtain several first fault causes and matching degrees.
5. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 4, characterized in that: The track circuit fault analysis training includes the following steps: Based on the first fault cause and abnormal data feature P, a feature case M is matched to the track circuit knowledge base on the server side. i The abnormal data feature P is included in the feature case M according to the weighted average ratio skew algorithm. i In the process, it completes self-learning training and automatic updates of the track circuit knowledge base on the server side; Based on the second fault cause and abnormal data feature P, if the second fault cause is already included in the track circuit knowledge base on the server side, then the feature case M in the track circuit knowledge base on the server side will be matched. i If the second cause of failure is not included in the track circuit knowledge base on the server side, a new feature case M is formed. i The outlier feature P is included in the feature case M according to the weighted average ratio skew algorithm. i In the process, it completes self-learning training and automatic updates of the track circuit knowledge base on the server side.
6. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 5, characterized in that: The weighted average ratio skew algorithm includes the following steps: Selecting outlier features P from t 1p With feature case M i Chinese 1i According to the expression: t 1i权重 =(t 1p +t 1i权重 *Number of samples) / (Number of samples + 1) gives M i Chinese 1i New weighted data; According to the expression: Sample number = Sample number + 1, we get the new sample number; Select t from the abnormal data features P in sequence 2p To t np , respectively corresponding to feature case M i t in 2i To t ni According to the above expression, we obtain the feature case M. i New weight data for each feature value.
7. A method for analyzing the electrical characteristics and diagnosing faults in a track circuit, applied in a server, characterized in that... Includes the following steps: Receive a download request from the host computer, wherein the download request is used to request the download of the track circuit knowledge base; Send receipt data to the host computer so that the host computer can update the track circuit knowledge base on the host computer based on the receipt data; The system receives abnormal data features and first information sent by the host computer. The first information includes the first fault cause and matching degree obtained by the host computer from the analysis of the abnormal case set in the knowledge base. The abnormal case set in the knowledge base is obtained by the host computer extracting abnormal data features and matching them with feature data in the track circuit knowledge base on the host computer. The abnormal data features are obtained by the host computer acquiring real-time electrical characteristic data of the track circuit operation sent by the slave computer and determining whether there is an abnormality. If the matching degree of the first fault cause is greater than the set threshold, then track circuit fault analysis training is performed and the track circuit knowledge base on the server is updated; otherwise, manual diagnosis is initiated. Receive abnormal data characteristics and second information sent by the host computer, wherein the second information includes a second fault cause obtained by manual diagnosis; If the second cause of failure is already included in the track circuit knowledge base on the server side, then track circuit failure analysis training is performed, and the track circuit knowledge base on the server side is updated; if the second cause of failure is not included in the track circuit knowledge base on the server side, then track circuit failure analysis training is performed to form new feature cases, and the track circuit knowledge base on the server side is updated.
8. The method for analyzing the electrical characteristics and diagnosing faults of track circuits in a section according to claim 7, characterized in that: The server comprises several servers, which merge the track circuit knowledge base based on the feature data and version information in the track circuit knowledge base on the server side.
9. A device for analyzing the electrical characteristics and diagnosing faults in a track circuit, characterized in that: It includes The download request sending module is used to send a download request to the server according to preset conditions, wherein the download request is used to request the download of the track circuit knowledge base; The receipt data receiving module receives receipt data from the server and updates the track circuit knowledge base on the host computer based on the receipt data. The real-time data acquisition module is used to acquire real-time electrical characteristic data of the track circuit operation sent by the lower-level machine; The anomaly detection module is used to determine whether there are any anomalies in the acquired real-time electrical characteristic data; The feature data matching module is used to extract abnormal data features when an anomaly exists, and match them with the data features in the track circuit knowledge base on the host computer to obtain a set of abnormal cases in the knowledge base that conform to the abnormal working state features. The first information sending module is used to analyze the abnormal case set of the knowledge base and obtain the first information, and send the abnormal data features and the first information to the server, so that the server can perform track circuit fault analysis training based on the abnormal data features and the first information and update the track circuit knowledge base on the server side. The first information includes the first fault cause and matching degree obtained by the host computer analysis. The track circuit fault analysis training is carried out when the matching degree of the first fault cause is greater than a set threshold. The second information sending module is used to send abnormal data features and second information obtained after manual diagnosis to the server, so that the server can perform track circuit fault analysis training based on the abnormal data features and second information and update the track circuit knowledge base on the server side. The second information includes the second fault cause obtained by manual diagnosis.
10. A device for analyzing the electrical characteristics and diagnosing faults in a track circuit, characterized in that: It includes A download request receiving module is used to receive download requests from a host computer, wherein the download request is used to request the download of a track circuit knowledge base; The receipt data sending module is used to send receipt data to the host computer so that the host computer can update the track circuit knowledge base on the host computer based on the receipt data. The first information receiving module is used to receive abnormal data features and first information sent by the host computer. The first information includes the first fault cause and matching degree obtained by the host computer from the analysis of the abnormal case set in the knowledge base. The abnormal case set in the knowledge base is obtained by the host computer extracting abnormal data features and matching them with feature data in the track circuit knowledge base on the host computer. The abnormal data features are obtained by the host computer acquiring real-time electrical characteristic data of the track circuit operation sent by the slave computer and judging whether there is an abnormality. The first information judgment module is used to determine whether the matching degree of the first fault cause in the first information is greater than a set threshold. If it is greater than the set threshold, track circuit fault analysis training is performed and the track circuit knowledge base on the server is updated; otherwise, manual diagnosis is initiated. The second information receiving module is used to receive abnormal data characteristics and second information sent by the host computer, wherein the second information includes the second fault cause obtained by manual diagnosis; The second information judgment module is used to determine whether the second fault cause is included in the track circuit knowledge base on the server side. If it is included, track circuit fault analysis training is performed and the track circuit knowledge base on the server side is updated; if it is not included, track circuit fault analysis training is performed to form new feature cases and the track circuit knowledge base on the server side is updated.
11. An electronic device, characterized in that: include At least one processor; A storage device for storing at least one computer program; When the at least one computer program is executed by the at least one processor, the at least one processor performs the method as described in any one of claims 1 to 8.
12. A computer-readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.