Ethernet-based Train Fault Monitoring and Prediction Method, System, and Storage Medium

Abstract

The present invention discloses a train fault monitoring and prediction method, system and storage medium based on Ethernet. The method includes: after receiving train operation data, the on-vehicle controller requests to establish a communication connection with an edge server corresponding to the target location where the train is located, and transmits the train operation data to the edge server; the edge server divides the train operation data into hot operation data and cold operation data, and requests to obtain historical operation data corresponding to the hot operation data for analysis based on a preset analysis model to generate a first analysis result for train fault monitoring; the cloud server searches for train parameters corresponding to the cold operation data, and analyzes the cold operation data and the train parameters to generate a second analysis result for train fault prediction. Through the analysis and processing of train operation data by Ethernet, the cloud-edge collaboration mechanism and a preset analysis model formed by training through a deep neural network, the present invention realizes timely and accurate monitoring and prediction of train faults.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a train fault monitoring and prediction method, system and storage medium based on Ethernet. Background Art

[0002] With the progress of science and technology, trains, especially high-speed trains, have developed into modern high-tech means of transportation. In order to ensure the safety, reliability, smoothness and comfort of train operation, it is necessary to collect various data of various devices and components during train operation to monitor and predict possible faults of the train.

[0003] Currently, the status data of various devices and components in the train is often transmitted and processed through the Train Communication Network (TCN). However, with the increase in intelligent devices and functions in the train, the types and volumes of data to be transmitted and analyzed are constantly increasing. The traditional train communication network built based on fieldbus has difficulty meeting the current transmission requirements and is also difficult to quickly and accurately process and analyze various types of data. As a result, it will lead to the inability to accurately and timely monitor and predict possible faults of the train, especially for data types with high real-time requirements, which will cause immeasurable losses.

[0004] Therefore, how to achieve timely and accurate monitoring and prediction of possible faults on the train is a technical problem that needs to be solved urgently at present. Summary of the Invention

[0005] The main purpose of the present invention is to provide a train fault monitoring and prediction method, system and storage medium based on Ethernet, aiming to solve the technical problem of how to achieve timely and accurate monitoring and prediction of possible faults on the train in the prior art.

[0006] To achieve the above object, the present invention provides a train fault monitoring and prediction method based on Ethernet. The train includes an on-vehicle controller and an on-vehicle PHM module communicatively connected to the on-vehicle controller through an Ethernet train backbone network. The train fault monitoring and prediction method includes:

[0007] After receiving the train operation data transmitted by the on-vehicle PHM module based on the Ethernet train backbone network, the on-vehicle controller determines the target location where the train is located and requests to establish a communication connection with an edge server corresponding to the target location to transmit the train operation data to the edge server;

[0008] The edge server divides the train operation data into hot operation data and cold operation data, requests the cloud server communicatively connected to the edge server to obtain historical operation data corresponding to the hot operation data, and transmits the cold operation data to the cloud server;

[0009] The edge server analyzes the hot operation data and the historical operation data based on a preset analysis model to generate a first analysis result, and monitors the train for faults according to the first analysis result;

[0010] The cloud server searches for train parameters corresponding to the cold operation data of the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict faults of the train.

[0011] Preferably, the hot operation data includes vibration signals corresponding to the motor shafts in the running gear of the train. The steps for the edge server to analyze the hot operation data and the train historical data based on a preset analysis model to generate a first analysis result include:

[0012] Perform time-domain analysis and frequency-domain analysis on the vibration signals respectively to obtain the original time-domain feature group and the original frequency-domain feature group of the vibration signals;

[0013] Filter the vibration signals, and extract a filtered time-domain feature group and a filtered frequency-domain feature group from the filtered vibration signals;

[0014] Form a feature matrix from the original time-domain feature group, the original frequency-domain feature group, the filtered time-domain feature group and the filtered frequency-domain feature group, and form a historical feature matrix from the historical operation data;

[0015] Perform fault analysis on the feature matrix and the historical feature matrix based on the input layer, the first hidden layer, the second hidden layer and the output layer in the preset analysis model to obtain the first analysis result, where the output function of the output layer is:

[0016]

[0017] where, a[x] represents the output value of the x-th node in the output layer of the preset analysis model, and g k 、g1、g2 respectively represent the activation functions of the output layer, the first hidden layer and the second hidden layer, T2 represents the number of nodes in the second hidden layer, w xj2 represents the third weight between the second hidden layer and the output layer, T1 represents the number of nodes in the first hidden layer, w j2j1 represents the second weight between the first hidden layer and the second hidden layer, T0 represents the number of nodes in the input layer, w j1iDenote the first weight between the input layer and the first hidden layer, p 1i Denote the i-th row of the feature matrix, p 2i Denote the i-th row of the historical feature matrix, B j1 、B j2 、B x Denote the bias values of the first hidden layer, the second hidden layer, and the output layer respectively.

[0018] Preferably, the first weight, the second weight, and the third weight are generated by a preset update formula group during the training process of the preset analysis model. The preset update formula group is as follows:

[0019]

[0020] Among them, w’ j1i Denote the updated first weight, w’ j2j1 Denote the updated second weight, w’ xj2 Denote the updated third weight, μ denotes the learning rate during the training process of the preset analysis model, p 3i Denote the i-th row of the training matrix during the training process of the preset analysis model, δ yj1 Is the training error rate of the y-th training of the preset analysis model on the first hidden layer, δ yj2 Is the training error rate of the y-th training of the preset analysis model on the second hidden layer, δ yx Is the training error rate of the y-th training of the preset analysis model on the output layer.

[0021] Preferably, the vehicle-mounted PHM module includes sub-modules corresponding to each carriage of the train respectively. Each sub-module transmits the operation data collected from each carriage to the node of the Ethernet train backbone network based on Ethernet networking, and is transmitted to the vehicle-mounted controller by the Ethernet train backbone network;

[0022] The step of performing fault monitoring on the train according to the first analysis result includes:

[0023] According to the node identifier carried in each vibration signal, find the target carriage corresponding to the first analysis result;

[0024] According to the target carriage, the fault type in the first analysis result, and the type parameter corresponding to the fault type, generate monitoring information for fault monitoring of the motor shaft of the running gear in the target carriage, and output the monitoring information.

[0025] Preferably, the cold operation data includes the image detection data and the sensing detection data of the wheel tread in the train. The step of analyzing the cold operation data and the train parameters to generate a second analysis result for fault prediction of the train includes:

[0026] Perform smoothing filtering and edge extraction on each depth image in the detected image data of the wheel tread to obtain multiple wheel tread edge maps, and convert each of the wheel tread edge maps into a binary edge map;

[0027] Generate a pixel matrix corresponding to each binary edge map according to the coordinate values of the pixels in the preset coordinate axes in each binary edge map, and perform a difference operation between each pixel matrix and the reference pixel matrix corresponding to the wheel tread to obtain a first operation matrix;

[0028] Perform a summation operation on each of the first operation matrices to obtain a second operation matrix, and generate a graph analysis result of the wheel tread according to the second operation matrix;

[0029] Generate a sensing analysis result of the wheel tread according to the vibration voltage value, instantaneous speed value in the sensing detection data, and the train parameters;

[0030] Generate a second analysis result according to the graph analysis result and the sensing analysis result to perform fault prediction on the wheel tread.

[0031] Preferably, the step of generating a sensing analysis result of the wheel tread according to the vibration voltage value, instantaneous speed value in the sensing detection data, and the train parameters includes:

[0032] Judge whether the vibration voltage value is greater than a preset threshold. If it is greater than the preset threshold, determine the operating section corresponding to the train wheel tread, and search for the section coefficient corresponding to the operating section, and the first instrument correction coefficient and the second instrument correction coefficient corresponding to the vibration voltage value and the instantaneous speed value respectively;

[0033] Perform calculations on the vibration voltage value, instantaneous speed value, section coefficient, and the wheel diameter and wheel average weight in the train parameters based on a preset calculation formula to obtain the calculation result. The preset calculation formula is:

[0034]

[0035] wherein, H represents the calculation result, D represents the wheel diameter, U represents the vibration voltage value, L represents the section coefficient, M represents the wheel average weight, V represents the instantaneous speed value, U0 represents the first instrument correction coefficient, and V0 represents the second instrument correction coefficient;

[0036] Compare the calculation result with the corresponding relationship between the preset bruise depth interval and the bruise grade to determine the bruise grade corresponding to the calculation result, and generate the sensing analysis result with the calculation result and the bruise grade.

[0037] Preferably, the step of generating the second analysis result based on the graph analysis result and the sensing analysis result to perform fault prediction on the wheel tread includes:

[0038] Determine whether both the graph analysis result and the sensing analysis result carry a bruise identification. If both carry a bruise identification, obtain the first bruise depth and the first bruise grade in the graph analysis result, and the second bruise depth and the second bruise grade in the sensing analysis result;

[0039] Generate a similarity value between the first bruise depth and the second bruise depth, and determine whether the similarity value is less than a preset similarity threshold, and whether the first bruise grade and the second bruise grade are the same;

[0040] If the similarity value is less than the preset similarity threshold, and the first bruise grade and the second bruise grade are the same, generate an average depth value between the first bruise depth and the second bruise depth, and generate the second analysis result based on the average depth value and the first bruise grade or the second bruise grade, and perform fault prediction on the wheel tread according to the second analysis result.

[0041] Preferably, the step of transmitting the cold operation data to the cloud server includes:

[0042] Search for the data to be redundant in the cold operation data and the corresponding target redundant data, and determine whether the data to be redundant is exactly the same as the target redundant data. If they are exactly the same, remove any one of the data to be redundant and the target redundant data from the cold operation data to obtain the data to be transmitted;

[0043] If the data to be redundant is not exactly the same as the target redundant data, form the data to be redundant and the target redundant data together as the data to be transmitted;

[0044] Generate an encryption key based on a preset encryption algorithm, and encrypt the data to be transmitted based on the encryption key to obtain encrypted data for transmission to the cloud server, where the preset encryption algorithm is:

[0045] Select any preset prime number from a preset prime number array, and generate a first random number and a first parameter pair;

[0046] Perform calculations on the preset prime number, the first random number, and the first parameter pair based on a first preset formula to obtain a first calculation result, and determine whether the first calculation result is a prime number. The first preset formula is:

[0047] y = a * [e h * x 2 + b;

[0048] Among them, y represents the first calculation result, a and b represent the first parameter pair, h represents the first random number, x represents a preset prime number, and [.] represents rounding down.

[0049] If the first calculation result is a prime number, generate a second random number and a second random number pair, and calculate the second random number, the second parameter pair, and the first calculation result based on the first preset formula to obtain a second calculation result.

[0050] Judge whether the second calculation result is a prime number. If it is a prime number, calculate the second calculation result based on the second preset formula to obtain a third calculation result.

[0051] Judge whether the third calculation result is a prime number. If it is a prime number, determine the third result as the first prime number.

[0052] Execute the step of screening any preset prime number from the preset prime number array and generating the first random number, generate a second prime number, and generate an encryption key according to the first prime number and the second prime number.

[0053] Furthermore, to achieve the above object, the present invention also provides an Ethernet-based train fault monitoring and prediction system. The Ethernet-based train fault monitoring and prediction system includes a storage, a processor, a communication bus, and a control program stored on the storage:

[0054] The communication bus is used to realize the connection communication between the processor and the storage.

[0055] The processor is used to execute the control program to realize the steps of the Ethernet-based train fault monitoring and prediction method as described above.

[0056] Furthermore, to achieve the above object, the present invention also provides a storage medium. A control program is stored on the storage medium. When the control program is executed by a processor, the steps of the Ethernet-based train fault monitoring and prediction method as described above are realized.

[0057] The present invention discloses a train fault monitoring and prediction method, system and storage medium based on Ethernet. The train is provided with an on-board controller and an on-board PHM module connected to the on-board controller through an Ethernet train backbone network. After receiving the train operation data transmitted by the on-board PHM module based on Ethernet, the on-board controller determines the target position of the train, and after requesting the edge server corresponding to the target position to establish a communication connection, transmits the train operation data to the edge server; the edge server divides the received train operation data into hot operation data and cold operation data, and initiates a request to obtain the historical operation data corresponding to the hot operation data to the cloud server connected to it, and transmits the cold operation data to the cloud server at the same time; after obtaining the historical operation data, the edge server analyzes the hot operation data and the historical operation data through a preset analysis model pre-set therein, generates a first analysis result, and then performs fault monitoring on the train according to the first analysis result; after receiving the cold operation data, the cloud server searches for the train parameters corresponding to the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict the fault of the train. In this way, by setting up a cloud-edge coordination mechanism between edge servers and cloud servers, hot running data with high real-time requirements are transmitted to edge servers near the train for processing, while cold running data with relatively low real-time requirements are transmitted to cloud servers for processing, so that data with high real-time requirements can be processed quickly and timely, avoiding the problem of inaccuracy caused by centralized data processing. In addition, the high transmission rate, short communication delay and high reliability of Ethernet further ensure the timeliness and accuracy of data processing. Timely and accurate monitoring and prediction of possible faults on the train are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flow chart of the first embodiment of the Ethernet-based train fault monitoring and prediction method of the present invention;

[0059] Figure 2 It is a flow chart of a second embodiment of the Ethernet-based train fault monitoring and prediction method of the present invention;

[0060] Figure 3 It is a flow chart of the third embodiment of the Ethernet-based train fault monitoring and prediction method of the present invention;

[0061] Figure 4 It is a structural schematic diagram of the hardware operating environment involved in an embodiment of the Ethernet-based train fault monitoring and prediction system of the present invention.

[0062] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0063] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] The present invention provides a method for monitoring and predicting train faults based on Ethernet. The train includes an on-vehicle controller and an on-vehicle PHM (Prognostics and Health Management) module communicatively connected to the on-vehicle controller through an Ethernet train backbone network. Please refer to Figure 1 , Figure 1 which is a schematic flowchart of the first embodiment of the method for monitoring and predicting train faults based on Ethernet according to the present invention.

[0065] The embodiments of the present invention provide an embodiment of the method for monitoring and predicting train faults based on Ethernet. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than here. Specifically, the method for monitoring and predicting train faults based on Ethernet in this embodiment includes:

[0066] Step S10, after receiving the train operation data transmitted by the on-vehicle PHM module based on the Ethernet train backbone network, the on-vehicle controller determines the target location where the train is located and requests to establish a communication connection with the edge server corresponding to the target location to transmit the train operation data to the edge server.

[0067] In this embodiment, the train can be a regular train, a bullet train, a high-speed train, or a subway, an intercity train, etc. Regardless of the type of train, a train communication network is provided in the train to provide public communication services for train operation control, multi-train coupling control, remote data transmission, monitoring and fault diagnosis of the working status of vehicle equipment, ground maintenance operations, crew services, and passenger services, etc., to realize train information transmission and sharing. In this embodiment, the train communication network is preferably formed by Ethernet and includes an Ethernet train backbone network (ETB) and an Ethernet consist network (ECN). The ETB can adopt a linear topology structure and connect each consist network through backbone network nodes to realize communication between each carriage. The ECN connects the terminal devices in the carriage and is responsible for data communication and network management between each node in the carriage or consist.

[0068] Further, on the train, there is an on-vehicle controller and an on-vehicle PHM module communicatively connected to the on-vehicle controller via an Ethernet train backbone network. The on-vehicle controller is used to regulate the overall state of all aspects of the train, while the on-vehicle PHM module is used to collect data related to train fault prediction and health management and cooperate with relevant servers to achieve fault prediction and health management. The on-vehicle PHM module is connected to various types of sensors via an Ethernet network, such as pressure sensors, temperature sensors, rotational speed sensors, vibration sensors, etc. During the train operation, various data collected by the sensors are transmitted to the on-vehicle PHM module via the Ethernet network, and the on-vehicle PHM module forms the train operation data from various data and transmits it to the on-vehicle controller via the Ethernet train backbone network.

[0069] Furthermore, due to the long-distance operation characteristics of the train, in order to perform fault prediction and health management on the long-distance running train more timely and quickly, in this embodiment, a server for realizing fault prediction and health management is constructed in a cloud-edge collaboration manner. Edge servers are set at different geographical locations to process data with high real-time requirements during the train operation, while a cloud server communicatively connected to each edge server is set at a certain location to process data with low real-time requirements during the train operation and store train parameters and various historical data. Among them, the location of the cloud server can be set at the central location of each edge server or the central location of the area with a large number of train operations.

[0070] Further, the train operation data includes train positioning data. The on-vehicle controller determines the target position of the train's travel through the train positioning data and searches for the edge server located at the target position or the edge server closest to the target position. Then the on-vehicle controller sends a request to establish a communication connection to the found edge server. After receiving the request, the edge server determines its own load and judges whether the amount of data it needs to process exceeds the amount of data it can bear. If it exceeds the amount of data it can bear, it returns to the on-vehicle controller the size of the current amount of data it needs to process and a prompt message about the waiting duration. The on-vehicle controller reselects a neighboring edge server based on the prompt message. The reselected edge server also performs a load judgment. If the load is less than the amount of data it can bear, it establishes a communication connection with the on-vehicle controller and returns a prompt message indicating that the communication connection has been successfully established to the on-vehicle controller. If the loads of multiple neighboring edge servers re-searched are all too high, the on-vehicle controller compares the waiting durations returned by each neighboring edge server and determines the neighboring edge server with the shortest waiting duration to send a request to establish a communication connection again. Then, after the communication connection is established, the train operation data is transmitted to the edge server for processing.

[0071] Step S20: The edge server divides the train operation data into hot operation data and cold operation data, requests the cloud server communicatively connected to the edge server to obtain the historical operation data corresponding to the hot operation data, and transmits the cold operation data to the cloud server.

[0072] Furthermore, an identifier is set in advance for the data with high real-time requirements in the train. The edge server screens the received train operation data according to the identifier and divides it into hot operation data and cold operation data. The hot operation data is the data with high real-time requirements, mainly the fault data affecting the real-time operation of the train, such as the data related to the running gear of the train; the cold operation data is the data with low real-time requirements, such as air-conditioning data, component wear data, etc. For the hot operation data, the edge cloud server processes it in combination with the previous operation data of the corresponding components. The previous operation data reflects the change of the operation state of the components, which can make the processing of the hot operation data more accurate. This type of operation data is stored in the cloud server as historical operation data. The edge server generates an acquisition request with the identifier of the hot operation data and sends it to the cloud server to request the historical operation data corresponding to the hot operation data.

[0073] For the cold operation data, the edge server transmits it to the cloud server for processing to prevent the edge server from occupying server resources due to processing the cold operation data and affecting the processing speed of the hot operation data. Moreover, to ensure the accuracy of data acquisition and transmission, redundant sensors are usually set for data acquisition, that is, two sensors of the same type are set to collect the same data. Thus, when both data acquisition and transmission are normal, the divided cold operation data contains two sets of the same data. When transmitting the cold operation data to the cloud server, it is necessary to screen the redundant data in the cold operation data to improve the transmission rate. Specifically, the step of transmitting the cold operation data to the cloud server includes:

[0074] Step S21: Search for the redundant data to be processed and the target redundant data corresponding to the redundant data to be processed in the cold operation data, and determine whether the redundant data to be processed is exactly the same as the target redundant data. If they are exactly the same, remove any one of the redundant data to be processed and the target redundant data from the cold operation data to obtain the data to be transmitted;

[0075] Step S22: If the redundant data to be processed is not exactly the same as the target redundant data, form the redundant data to be processed and the target redundant data together as the data to be transmitted;

[0076] Step S23: Generate an encryption key based on a preset encryption algorithm, and encrypt the data to be transmitted based on the encryption key to obtain encrypted data for transmission to the cloud server, where the preset encryption algorithm is as follows:

[0077] Step S24: Select any preset prime number from a preset prime number array, and generate a first random number and a first parameter pair;

[0078] Step S25: Calculate the preset prime number, the first random number, and the first parameter pair based on a first preset formula to obtain a first calculation result, and determine whether the first calculation result is a prime number;

[0079] Step S26: If the first calculation result is a prime number, generate a second random number and a second random number pair, and calculate the second random number, the second parameter pair, and the first calculation result based on the first preset formula to obtain a second calculation result;

[0080] Step S27: Determine whether the second calculation result is a prime number. If it is a prime number, calculate the second calculation result based on a second preset formula to obtain a third calculation result;

[0081] Step S28: Determine whether the third calculation result is a prime number. If it is a prime number, determine the third result as the first prime number;

[0082] Step S29: Execute the step of selecting any preset prime number from the preset prime number array and generating a first random number to generate a second prime number, and generate an encryption key based on the first prime number and the second prime number.

[0083] Further, the components that need to be redundantly set are used as the devices to be redundantly set, and the data they collect is correspondingly the data to be redundantly set, while the components that perform redundancy on the components that need to be redundant are the target redundant devices, and the data they collect is correspondingly the target redundant data. Moreover, the device identifiers of the devices to be redundantly set and the target redundant devices are carried between the data to be redundantly set and the target redundant data, indicating the redundancy relationship between the data to be redundantly set and the target redundant data and the devices from which they originate. Before the edge server transmits the cold running data to the cloud server, it first searches for the data to be redundantly set in the cold running data and the corresponding target redundant data according to the device identifiers carried in each data in the cold running data. Then, the data to be redundantly set and the target redundant data are compared to determine whether the data to be redundantly set and the target redundant data are exactly the same. If they are exactly the same, it means that the data collected by the devices to be redundantly set and the target redundant devices is the same, and any one of them can be selected for transmission to the cloud server. At this time, the data to be redundantly set or the target redundant data can be removed from the cold running data, and the remaining data is used as the data to be transmitted and transmitted to the cloud server for processing.

[0084] Conversely, if it is determined through comparison that the data to be redundant and the target redundant data are not exactly the same, it indicates that there may be incorrect data detected by one of the devices to be redundant and the target redundant device. At this time, the data to be redundant and the target redundant data are jointly formed into the data to be transmitted and transmitted to the cloud server, and the cloud server makes a judgment in combination with the previous detection data. The data that differs little from the previous detection data is selected from the data to be redundant and the target redundancy as valid data for processing.

[0085] In addition, to ensure the security of the transmitted data, the edge server also encrypts the data to be transmitted. An encryption algorithm is preset, and an encryption key is generated through the preset encryption algorithm. Then, the data to be transmitted is encrypted through the encryption key to obtain encrypted data and transmitted to the cloud server. The preset encryption algorithm is preferably an asymmetric encryption algorithm. Specifically, a preset prime number array is preset, and to ensure security, the values of the data in the preset prime number array can be set to be relatively large. For example, the prime numbers between 100 and 200 are formed into a preset prime number array. Any preset prime number is selected from the preset prime number array, and three random numbers are generated through a random number generator. One of them is selected as the first random number, and the remaining two are used as the first parameter pair. Then, according to the first preset formula, calculations are performed on the preset prime number, the first random number, and the first parameter pair to obtain a first calculation result. It is judged whether the first calculation result is a prime number. If it is not a prime number, a random number is generated again through the random number generator to update the first random number, and calculations are performed with the first parameter pair and the preset prime number to obtain the first calculation result again for judgment until the obtained first calculation result is a prime number. At this time, three random numbers are generated again through the random number generator. One of them is selected as the second random number, and the remaining two are used as the second parameter pair. According to the first preset formula, calculations are performed on the first calculation result, the second random number, and the second parameter pair to obtain a second calculation result, and it is judged whether the second calculation result is a prime number. If it is not a prime number, a random number is generated through the random number generator to update the second random number, and calculations are performed with the first calculation result and the second parameter pair to obtain the second calculation result again for judgment until the obtained second calculation result is a prime number. The first preset formula is specifically shown in the following formula (1).

[0086] y=a*[e h *x 2 +b (1);

[0087] Among them, corresponding to step S25, y represents the first calculation result, a and b represent the first parameter pair, h represents the first random number, and x represents a preset prime number; corresponding to step S26, y represents the second calculation result, a and b represent the second parameter pair, h represents the second random number, x represents the first calculation result, and [.] represents rounding down. Through the differences between the first random number and the second random number, as well as the differences between the first parameter pair and the second parameter pair, the parameters of the first preset formula are transformed, so that the first calculation result and the second calculation result are generated with different parameters, which is beneficial to improving the security of the encryption algorithm.

[0088] Furthermore, the second calculation result is calculated through a second preset formula to determine whether the third calculation result is a prime number. If it is not a prime number, two random numbers are regenerated according to the random number generator to update the second parameter pair, and based on the updated second parameter pair, the second calculation result and the third calculation result are generated in sequence for judgment until the obtained third calculation result is a prime number, and the third calculation result is used as the first prime number. After that, any preset prime number is selected again from the preset prime number array to generate the second prime number, and the encryption and decryption keys are generated based on the first prime number and the second prime number. Among them, the encryption and decryption keys include the public key as the encryption key and the private key as the decryption key. The edge server encrypts the data to be transmitted through the encryption key and then transmits it to the cloud server, and the cloud server decrypts it through the decryption key. Since the first prime number and the second prime number are generated based on the prime numbers calculated by the preset prime number array, the first preset formula, and the second preset formula, and the prime numbers in the preset prime number array are relatively large, combined with the parameter transformation calculation of the first preset formula, the values of the generated first prime number and the second prime number can also be relatively large, and the generation process is more complex, thereby increasing the complexity of encryption and decryption and making data transmission more secure.

[0089] Step S30, the edge server analyzes the hot operation data and the historical operation data based on a preset analysis model to generate a first analysis result, and monitors the faults of the train according to the first analysis result.

[0090] Furthermore, for various types of hot operation data to be processed, a preset analysis model is pre-constructed and trained in the edge server through a neural network. Through this preset analysis model, the hot operation data and the historical operation data are analyzed to generate a first analysis result. The first analysis result includes information such as whether there is a fault in the components corresponding to the hot operation data, the carriage where the fault point is located, the severity of the fault, and the possible reasons if there is a fault. Corresponding monitoring information is generated through this type of information in the first analysis result for alarm, realizing the fault monitoring of the train.

[0091] Step S40, the cloud server searches for train parameters corresponding to the cold operation data of the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict faults of the train.

[0092] Furthermore, for the cold operation data, the cloud server searches for the train parameters corresponding to the cold operation data. For example, if the cold operation data is air conditioning data, the train parameters to be searched are the model and power of the air conditioner, and if the cold operation data is wheel tread data, the train parameters to be searched are the diameter and weight of the wheel, etc. The fixed parameters of the components corresponding to the cold operation data in the train are used to reflect their influence on the cold operation data. After finding the corresponding train parameters, the cloud server analyzes the train parameters and the cold operation data together to obtain a second analysis result. The second analysis result includes information such as whether the components corresponding to the cold operation data may fail and the probability of failure. The corresponding warning information is generated through this type of information in the second analysis result to predict train failures.

[0093] The Ethernet-based train fault monitoring and prediction method of this embodiment is provided with an on-board controller on the train, and an on-board PHM module connected to the on-board controller through the Ethernet train backbone network. After receiving the train operation data transmitted by the on-board PHM module based on Ethernet, the on-board controller determines the target position of the train, and after requesting the edge server corresponding to the target position to establish a communication connection, transmits the train operation data to the edge server; the edge server divides the received train operation data into hot operation data and cold operation data, and initiates a request to obtain the historical operation data corresponding to the hot operation data to the cloud server connected to it, and transmits the cold operation data to the cloud server at the same time; after obtaining the historical operation data, the edge server analyzes the hot operation data and the historical operation data through the preset analysis model pre-set therein, generates a first analysis result, and then performs fault monitoring on the train according to the first analysis result; after receiving the cold operation data, the cloud server searches for the train parameters corresponding to the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict the train fault. In this way, by setting up a cloud-edge coordination mechanism between edge servers and cloud servers, hot running data with high real-time requirements are transmitted to edge servers near the train for processing, while cold running data with relatively low real-time requirements are transmitted to cloud servers for processing, so that data with high real-time requirements can be processed quickly and timely, avoiding the problem of inaccuracy caused by centralized data processing. In addition, the high transmission rate, short communication delay and high reliability of Ethernet further ensure the timeliness and accuracy of data processing. Timely and accurate monitoring and prediction of possible faults on the train are achieved.

[0094] Further, please refer to Figure 2 , based on the first embodiment of the train fault monitoring and prediction method based on Ethernet of the present invention, a second embodiment of the train fault monitoring and prediction method based on Ethernet of the present invention is proposed.

[0095] The difference between the second embodiment of the train fault monitoring and prediction method based on Ethernet and the first embodiment of the train fault monitoring and prediction method based on Ethernet is that the hot operation data includes vibration signals corresponding to the motor shafts in the running gear of the train, and the steps for the edge server to analyze the hot operation data and the train historical data based on a preset analysis model to generate a first analysis result include:

[0096] Step S31, perform time-domain analysis and frequency-domain analysis on the vibration signals respectively to obtain an original time-domain feature group and an original frequency-domain feature group of the vibration signals;

[0097] Step S32, filter the vibration signals, and extract a filtered time-domain feature group and a filtered frequency-domain feature group from the filtered vibration signals;

[0098] Step S33, form a feature matrix with the original time-domain feature group, the original frequency-domain feature group, the filtered time-domain feature group and the filtered frequency-domain feature group, and form a historical feature matrix with the historical operation data;

[0099] Step S34, perform fault analysis on the feature matrix and the historical feature matrix based on the input layer, the first hidden layer, the second hidden layer and the output layer in the preset analysis model to obtain the first analysis result.

[0100] Further, each carriage of the train includes a running gear for walking, and the running gear includes a motor shaft. The hot operation data in this embodiment includes vibration signals corresponding to such motor shafts. The edge server performs time-domain analysis and frequency-domain analysis on such vibration signals. Through time-domain analysis, multiple time-domain features are obtained and formed into an original time-domain feature group of the vibration signals, which at least includes skewness, kurtosis, peak factor, margin factor, waveform factor, pulse factor, etc.; at the same time, through frequency-domain analysis, multiple frequency-domain features are obtained and formed into an original frequency-domain feature group of the vibration signals, which at least includes: spectrum, energy spectrum, power spectrum, frequency band width, power spectrum, etc. At the same time, the vibration signals are filtered, and time-series analysis and frequency-domain analysis are performed on the filtered vibration signals. Through the analysis, corresponding time-domain features and frequency-domain features are extracted and formed into a filtered time-domain feature group and a filtered frequency-domain feature group.

[0101] Further, the original time-domain feature group, the original frequency-domain feature group, the filtered time-domain feature group, and the filtered frequency-domain feature group are jointly formed into a feature matrix. One feature group is one row in the feature matrix. For the case where the number of elements in the feature groups is different, the feature group lacking elements is supplemented with the value 1 so that each matrix row in the feature matrix has the same number of elements. Moreover, the historical operation data is the average value of the above-mentioned time-domain features obtained from the recent several times of fault monitoring of the motor shafts of the running gear of each carriage, and the average value of the above-mentioned frequency-domain features obtained. Similarly, the average value of each time-domain feature is formed into a historical time-domain feature group, and the average value of each frequency-domain feature is formed into a historical frequency-domain feature group. Furthermore, each feature group is formed into a historical feature matrix.

[0102] Further, the preset analysis model in the edge server includes an input layer, a hidden layer, and an output layer. Moreover, in order to improve the model analysis effect while avoiding model overfitting, the hidden layer is set to two layers, namely the first hidden layer and the second hidden layer. Fault analysis is performed on the feature matrix and the historical feature matrix through the input layer, the first hidden layer, the second hidden layer, and the output layer in the preset analysis model to obtain a first analysis result. Among them, the output function expression of the output layer is shown in the following formula (2).

[0103]

[0104] Among them, a[x] represents the output value of the x-th node in the output layer of the preset analysis model, and g k , g1, and g2 respectively represent the activation functions of the output layer, the first hidden layer, and the second hidden layer. T2 represents the number of nodes in the second hidden layer, and w xj2 represents the third weight between the second hidden layer and the output layer. T1 represents the number of nodes in the first hidden layer, and w j2j1 represents the second weight between the first hidden layer and the second hidden layer. T0 represents the number of nodes in the input layer, and w j1i represents the first weight between the input layer and the first hidden layer. p 1i represents the i-th row of the feature matrix, and p 2i represents the i-th row of the historical feature matrix. B j1 , B j2 , and Bx respectively represent the bias values of the first hidden layer, the second hidden layer, and the output layer. By forming the data obtained from the recent several times of fault monitoring of the motor shafts of the running gear of each carriage into a historical feature matrix, the change trend of the vibration signal of the motor shaft reflected by the historical feature matrix enables the first analysis result obtained through the analysis of the preset analysis model to more accurately reflect the current operating state of the motor shaft.

[0105] Moreover, to facilitate the determination of the carriage where the fault point is located, the on-vehicle PHM module can be divided into sub-modules corresponding to each carriage of the train. Each of the sub-modules transmits the operation data of each carriage collected based on Ethernet networking to the nodes of the Ethernet train backbone network, and is transmitted by the Ethernet train backbone network to the on-vehicle controller. The operation data transmitted by each sub-module constitutes the train operation data, including the hot operation data and cold operation data corresponding to each carriage. The first analysis result generated based on the vibration data of the motor shaft of the running gear of each carriage also reflects the fault condition of the motor shaft of the running gear of each carriage. One vibration data corresponds to one carriage, and one first analysis result is generated. Therefore, the fault monitoring of the motor shaft of the running gear of each carriage can be carried out based on each first analysis result. Specifically, the step of performing fault monitoring on the train according to the first analysis result includes:

[0106] Step S35: According to the node identifier carried by each vibration signal, find the target carriage corresponding to the first analysis result;

[0107] Step S36: Generate monitoring information for fault monitoring of the motor shaft of the running gear in the target carriage according to the target carriage, the fault type in the first analysis result, and the type parameter corresponding to the fault type, and output the monitoring information.

[0108] Furthermore, each carriage is provided with its own corresponding sub-module, and each sub-module is connected to the nodes of the Ethernet train backbone network. Therefore, one carriage corresponds to one node, and each carriage can be distinguished by numbering the nodes. The operation data from the same carriage all carry the node identifier corresponding to that carriage. Similarly, for the vibration signals of the motor shafts of the running gears of each carriage, they also carry the node identifiers of each carriage. After generating the first analysis result based on the vibration signal, determine the carriage where the vibration signal comes from through the node identifier carried by the vibration signal, and use this carriage as the target carriage corresponding to the first analysis result.

[0109] Further, the first analysis result includes information indicating whether there is a fault in the motor shaft. If this information indicates that there is no fault in the motor shaft, the target carriage and this type of information are jointly generated into monitoring information indicating no fault, which is used to indicate that there is no fault in the motor shaft of the running gear of this carriage. If this information indicates that there is a fault in the motor shaft, the fault type in the first analysis result, the type parameter corresponding to this fault type, and the target carriage are jointly generated into monitoring information and output, which is used to monitor the fault of the motor shaft in the running gear of the target carriage through this monitoring information. Among them, the fault type at least includes any one of journal wear, keyway wear, insufficient lubrication, pitting fault, and scratch fault, and the type parameter corresponding to the fault type at least includes a fault severity parameter, such as the wear degree, and the specific location of the fault, such as the specific location of pitting, etc. By jointly generating this type of information into monitoring information, it is convenient to quickly troubleshoot faults.

[0110] It should be noted that the preset analysis model is pre-trained through a large number of training samples. The training samples are data containing faults such as journal wear, keyway wear, insufficient lubrication, pitting fault, and scratch fault and marked with the fault severity, so that the preset analysis model can analyze the fault type and fault parameters to obtain the fault type and type parameter in the first analysis result. Moreover, in order to improve the accuracy of the fault analysis of the preset analysis model, during the training process of the preset analysis model, the first weight between the input layer and the first hidden layer, the second weight between the first hidden layer and the second hidden layer, and the third weight between the second hidden layer and the output layer are updated according to the preset update formula group. The specific preset update formula group can be seen in the following formula (3).

[0111]

[0112] Among them, w′ j1i represents the updated first weight, W′ j2j1 represents the updated second weight, W′ xj2 represents the updated third weight, μ represents the learning rate during the training process of the preset analysis model, p 3i represents the i-th row of the training matrix during the training process of the preset analysis model, δ yj1 is the training error rate of the y-th training of the preset analysis model on the first hidden layer, δ yj2 is the training error rate of the y-th training of the preset analysis model on the second hidden layer, δ yx is the training error rate of the y-th training of the preset analysis model on the output layer. The meanings of other symbols are the same as those represented by the above formula (2) and will not be elaborated here.

[0113] An iteration count is preset. When the number of training times for the preset analysis model reaches this iteration count, the preset analysis model is tested to generate a total error, and it is determined whether the total error is less than a preset threshold. If it is less than the preset threshold, it indicates that the accuracy of the preset analysis model meets the requirements, and the training of the preset analysis model is completed. If the total error is greater than the preset threshold, it indicates that the accuracy of the preset analysis model is relatively low and iterative training needs to continue. At this time, the total error is decomposed into the training error rates of each layer of the first hidden layer, the second hidden layer, and the output, and the first weight, the second weight, and the third weight are updated according to the training error rates of each layer through a preset update formula group until the total error generated by testing the preset analysis model is less than the preset threshold.

[0114] In this embodiment, time-domain features and frequency-domain features are extracted from the hot operating data, and combined with the historical features generated by the previous several fault detections, and analyzed by a preset analysis model pre-trained based on a neural network. On the one hand, the time-domain features and frequency-domain features can accurately reflect the characteristics of the hot operating data itself. On the other hand, the historical features reflect the change trend of the hot operating data at different fault detection times. Moreover, the weights of the preset analysis model are updated through the error rates of each layer in each training, ensuring the accuracy of the preset analysis model. In this way, the accuracy of hot operating data analysis is improved from multiple aspects.

[0115] Further, please refer to Figure 3 , based on the first and second embodiments of the train fault monitoring and prediction method based on Ethernet of the present invention, a third embodiment of the train fault monitoring and prediction method based on Ethernet of the present invention is proposed.

[0116] The difference between the third embodiment of the train fault monitoring and prediction method based on Ethernet and the first and second embodiments of the train fault monitoring and prediction method based on Ethernet is that the cold operating data includes the image detection data and the sensing detection data of the wheel tread in the train, and the step of analyzing the cold operating data and the train parameters to generate a second analysis result for fault prediction of the train includes:

[0117] Step S41, perform smoothing filtering and edge extraction processing on each depth image in the image detection data to obtain a plurality of tread edge images, and convert each of the tread edge images into a binary edge image;

[0118] Step S42, generate a pixel matrix corresponding to each binary edge image according to the coordinate values of the pixels in each binary edge image in a preset coordinate axis, and perform a difference operation between each pixel matrix and the reference pixel matrix corresponding to the wheel tread to obtain a first operation matrix;

[0119] Step S43: Perform a summation operation on each of the first operation matrices to obtain a second operation matrix, and generate a graphical analysis result of the wheel tread based on the second operation matrix;

[0120] Step S44: Generate a sensing analysis result of the wheel tread based on the vibration voltage value, the instantaneous speed value in the sensing detection data, and the train parameters;

[0121] Understandably, when the train runs along the track, the part where the train wheel contacts the rail is called the tread. As the train runs, friction may occur when the wheel tread contacts the rail, and thus wheel tread damage may be caused by the friction. Although wheel tread damage does not immediately affect the operation of the train, its impact cannot be underestimated in the long run. When the wheel tread is damaged, vibrations will be caused during the train operation, which will not only damage the train parts but also damage the rail. Moreover, the greater the depth of the wheel tread damage, the greater the vibration, reducing the running stability of the train. This embodiment is to predict the faults of the wheel tread.

[0122] Specifically, the cold operation data collected for the wheel tread includes graphical detection data and sensing detection data. The graphical detection data is to collect images of the wheel tread through a binocular vision camera, and the sensing detection data is to detect data related to the wheel tread through detection devices such as a voltmeter and a speed sensor. And the graphical detection data includes depth images of the wheel tread at multiple angles. After performing smoothing filtering and edge extraction processing on the multiple depth images, multiple tread edge maps containing the tread edge contours are obtained. Then, each tread edge map is converted into a black-and-white binary image to obtain multiple binary edge maps.

[0123] Further, a preset coordinate axis is pre-established. The preset coordinate axis can be an image coordinate axis corresponding to each binary edge map, with the upper left corner point of the image as the coordinate origin, the horizontal direction of the image as the X-axis, and the vertical direction of the image as the Y-axis. Then, for each binary edge map, obtain the coordinate values of each pixel in the preset coordinate axis, and generate a pixel matrix of each binary edge map according to the x coordinate and y coordinate positions corresponding to each pixel. Each element in the pixel matrix corresponds to the coordinate value of a pixel point.

[0124] Further, before the wheel tread is put into use, its image is collected to generate a reference pixel matrix corresponding to the wheel tread. The difference operation is performed between each pixel matrix and the reference pixel matrix to obtain a first operation matrix. For a wheel tread without abrasion, the value of each element in the first operation matrix obtained through the difference operation should be zero. However, if the value of a certain element is not zero, it indicates that the wheel tread has abrasion, causing the pixel points of its imaging to change. Therefore, the sum of the first operation means can be calculated to obtain a second operation matrix, and the elements with non-zero values in the second operation matrix are determined. The non-zero element reflects the abrasion, and the number and position of the non-zero elements reflect the severity and position depth of the abrasion. Therefore, the graph analysis result of the wheel tread abrasion condition can be generated based on the second operation matrix.

[0125] Furthermore, for the sensing detection data, the abrasion condition of the wheel tread is analyzed from the perspective of sensor detection, which includes vibration voltage values and instantaneous speed values. At the same time, in combination with train parameters, a sensing analysis result of the wheel tread is generated. Specifically, the step of generating the sensing analysis result of the wheel tread according to the vibration voltage value, instantaneous speed value in the sensing detection data, and the train parameters includes:

[0126] Step S441: Determine whether the vibration voltage value is greater than a preset threshold. If it is greater than the preset threshold, determine the operating section corresponding to the train tread, and find the section coefficient corresponding to the operating section, as well as the first instrument correction coefficient and the second instrument correction coefficient corresponding to the vibration voltage value and the instantaneous speed value respectively;

[0127] Step S442: Calculate the vibration voltage value, instantaneous speed value, section coefficient, and the wheel diameter and wheel average weight in the train parameters based on a preset calculation formula to obtain the calculation result;

[0128] Step S443: Compare the calculation result with the corresponding relationship between the preset abrasion depth interval and abrasion grade, determine the abrasion grade corresponding to the calculation result, and generate the sensing analysis result with the calculation result and the abrasion grade.

[0129] Furthermore, considering the influence of train design and manufacturing processes, after a period of operation, there will be a certain degree of abrasion on the train. Slight abrasion will not affect the train operation. Therefore, in this embodiment, a preset threshold is set in advance. The vibration voltage value in the sensing detection data is compared with the preset threshold to determine whether the vibration voltage value is greater than the preset threshold. If it is not greater than the preset threshold, it indicates that the abrasion is slight and no treatment is required. Conversely, if the vibration voltage value is greater than the preset threshold, the operating section corresponding to the train tread is determined. Different operating sections have different road conditions and different degrees of influence on abrasion. Corresponding section coefficients are set in advance for different sections. After determining the corresponding operating section, the section coefficient corresponding to this operating section is found. Then, according to the preset calculation formula, the vibration voltage value, the instantaneous speed value, the found section coefficient, and the train parameters including the wheel diameter and the average wheel weight are calculated to obtain a calculation result, which is the abrasion depth. The preset calculation formula for calculation can be seen in the following formula (4).

[0130]

[0131] Among them, H represents the calculation result, D represents the wheel diameter, U represents the vibration voltage value, L represents the section coefficient, M represents the average wheel weight, V represents the instantaneous speed value, U0 represents the first instrument correction coefficient, and V0 represents the second instrument correction coefficient. And formula (4) is the relationship expression between the abrasion depth and the vibration voltage, instantaneous speed, average wheel weight, and wheel diameter obtained through a large number of test data. The instrument errors of the instruments for detecting the vibration voltage value and the instantaneous speed value are represented by the instrument correction coefficients, and formula (4) is corrected, so as to fully consider various factors affecting the abrasion depth and make the calculation of the abrasion depth more accurate.

[0132] Further, in order to represent the severity of the wheel tread abrasion, a corresponding relationship between the abrasion depth interval and the abrasion grade is set in advance. The calculated result is compared with the abrasion depth interval in this corresponding relationship to determine the abrasion interval where the calculation result is located. Then, the abrasion grade corresponding to this abrasion depth interval in the corresponding relationship is found as the abrasion grade corresponding to the calculation result. Then, the determined abrasion grade and the calculation result representing the abrasion depth are jointly formed into a sensing analysis result. Among them, the larger the value of the calculation result, the deeper the abrasion, and the higher the corresponding abrasion grade, indicating the more serious the abrasion of the wheel tread.

[0133] Step S45, generate a second analysis result based on the graph analysis result and the sensing analysis result to perform a fault prediction on the wheel tread.

[0134] Furthermore, after the graph detection data and the sensing detection data respectively generate their own graph analysis results and sensing analysis results, the second analysis result can be generated by combining the difference between the two, and the train can be fault predicted based on the second analysis result. Specifically, the steps of generating the second analysis result for fault prediction of the train tread according to the graph analysis result and the sensing analysis result include:

[0135] Step S451, determine whether both the graph analysis result and the sensing analysis result carry a bruise mark. If both carry a bruise mark, obtain the first bruise depth and the first bruise grade in the graph analysis result, and the second bruise depth and the second bruise grade in the sensing analysis result;

[0136] Step S452, generate a similarity value between the first bruise depth and the second bruise depth, and determine whether the similarity value is less than a preset similarity threshold, and whether the first bruise grade and the second bruise grade are the same;

[0137] Step S453, if the similarity value is less than the preset similarity threshold, and the first bruise grade and the second bruise grade are the same, generate an average depth value between the first bruise depth and the second bruise depth, and generate the second analysis result based on the average depth value and the first bruise grade or the second bruise grade, and perform fault prediction on the wheel tread according to the second analysis result.

[0138] It can be understood that the graph analysis result generated based on the graph detection data and the sensing analysis result generated based on the sensing detection data are essentially used to judge whether there is a bruise on the wheel tread and the severity of the bruise from the graph dimension and the sensing dimension, so as to avoid the error of a single detection method and the situation of detection equipment failure. When both the graph detection and the sensing detection are correct, the detection results of the two should be relatively close. Therefore, the bruise fault of the wheel tread can be predicted based on the difference between the two.

[0139] Specifically, determine whether both the graph analysis result and the sensing analysis result carry a bruise mark. If both carry a bruise mark, it means that both the graph detection and the sensing detection determine that there is a bruise on the wheel tread, and it is necessary to further judge the severity of the bruise. If neither the graph analysis result nor the sensing analysis result carries a bruise mark, it means that both the graph detection and the sensing detection determine that there is no bruise on the wheel tread, and no further processing is required. If one of the graph analysis result and the sensing analysis result carries a bruise mark and the other does not, it means that the detection error between the graph detection and the sensing detection is relatively large. At this time, it can be analyzed according to the previous detection situation, or a prompt message for repeated detection can be output.

[0140] Further, for the case where abrasion marks are carried, the cloud server obtains the corresponding abrasion depth and abrasion grade in the graph analysis result as the first abrasion depth and the first abrasion grade. At the same time, it also obtains the abrasion depth and abrasion grade in the sensing analysis result as the second abrasion depth and the second abrasion grade. Then, a similarity calculation is performed between the first abrasion depth and the second abrasion depth to generate a similarity value therebetween. To standardize the proximity between the two, a preset similarity threshold is set in advance. The calculated similarity value is compared with the preset similarity threshold to determine whether the similarity value is less than the preset similarity threshold. If it is less than the preset similarity threshold, it indicates that the proximity between the first abrasion depth and the second abrasion depth is high; otherwise, it indicates that the difference between the first abrasion depth and the second abrasion depth is large. At the same time, it is also necessary to determine whether the first abrasion grade is the same as the second abrasion grade.

[0141] Furthermore, if it is determined that the similarity value is less than the preset similarity threshold and the first abrasion grade is the same as the second abrasion grade, it indicates that the detection results of the graph detection and the sensing detection are basically the same and are both accurate results. At this time, the first abrasion depth and the second abrasion depth are generated into an average depth value, and any one of the first abrasion grade and the second abrasion grade is selected to jointly generate a second analysis result for fault prediction of the train with the average depth value.

[0142] Further, if it is determined that the similarity value is greater than or equal to the preset similarity threshold and the first abrasion grade and the second abrasion grade are the same, it indicates that the graph detection and the sensing detection are consistent in the judgment of the abrasion grade but have a large difference in the determination of the abrasion depth. At this time, the first abrasion depth and the second abrasion depth are compared to determine the larger value between the two. Then, based on this larger value, combined with the first abrasion grade or the second abrasion grade, a second analysis result for fault prediction of the train is generated. Generating the second analysis result for abrasion fault prediction by selecting the larger value between the first abrasion depth and the second abrasion depth is beneficial to ensuring the smooth operation of the train.

[0143] Furthermore, if it is determined that the similarity value is greater than or equal to the preset similarity threshold and the first abrasion grade and the second abrasion grade are not the same, it indicates that the graph detection and the sensing detection are inconsistent in the judgment of the abrasion grade and also have a large difference in the determination of the abrasion depth. At this time, the first abrasion depth and the second abrasion depth are compared to determine the larger value between the two, and the higher grade among the first abrasion grade and the second abrasion grade is selected. Then, based on this larger value and the higher grade, prediction information for fault prediction of the train is generated. Generating the second analysis result for abrasion fault prediction by selecting the larger value between the first abrasion depth and the second abrasion depth and the higher grade is beneficial to ensuring the smooth operation of the train.

[0144] In this embodiment, a detection method is provided in which the detection of the wheel tread setting diagram and the sensing detection are arranged in parallel. By analyzing the detection data of the two detection methods, respective corresponding detection and analysis results are generated, and a second analysis result for predicting the wheel tread bruise fault is generated from the differences between the respective detection and analysis results of the two, ensuring the accuracy of the prediction of the wheel tread bruise fault.

[0145] In addition, an embodiment of the present invention also provides a train fault monitoring and prediction system based on Ethernet. Please refer to Figure 4 , Figure 4 which is a schematic structural diagram of the hardware operating environment of the equipment involved in the embodiment solution of the train fault monitoring and prediction system based on Ethernet of the present invention.

[0146] As Figure 4 shown, the train fault monitoring and prediction system based on Ethernet may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard). Optionally, the user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0147] Those skilled in the art can understand that Figure 4 the hardware structure of the train fault monitoring and prediction system based on Ethernet shown in

[0148] does not constitute a limitation on the train fault monitoring and prediction system based on Ethernet, and may include more or fewer components than shown in the figure, or combine some components, or arrange different components. Figure 4 shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a control program. Among them, the operating system is a program for managing and controlling the train fault monitoring and prediction system based on Ethernet and software resources, and supports the operation of the network communication module, the user interface module, the control program, and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0149] In Figure 4In the hardware structure of the Ethernet-based train fault monitoring and prediction system shown, the network interface 1004 is mainly used to connect to other system servers and communicate data with other system servers; the user interface 1003 is mainly used to connect to the client (user side) and communicate data with the client; the processor 1001 can call the control program stored in the storage 1005 and perform the following operations:

[0150] After receiving the train operation data transmitted by the on-vehicle PHM module based on the Ethernet train backbone network, the on-vehicle controller determines the target location where the train is located and requests to establish a communication connection with the edge server corresponding to the target location to transmit the train operation data to the edge server;

[0151] The edge server divides the train operation data into hot operation data and cold operation data, requests the cloud server communicating with the edge server to obtain the historical operation data corresponding to the hot operation data, and transmits the cold operation data to the cloud server;

[0152] The edge server analyzes the hot operation data and the historical operation data based on a preset analysis model, generates a first analysis result, and monitors the train for faults according to the first analysis result;

[0153] The cloud server searches for the train parameters corresponding to the cold operation data of the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict the faults of the train.

[0154] Further, the hot operation data includes vibration signals corresponding to the motor shafts in the running gear of the train. The steps for the edge server to analyze the hot operation data and the train historical data based on a preset analysis model to generate a first analysis result include:

[0155] Perform time-domain analysis and frequency-domain analysis on the vibration signals respectively to obtain the original time-domain feature group and the original frequency-domain feature group of the vibration signals;

[0156] Filter the vibration signals and extract the filtered time-domain feature group and the filtered frequency-domain feature group from the filtered vibration signals;

[0157] Form a feature matrix with the original time-domain feature group, the original frequency-domain feature group, the filtered time-domain feature group and the filtered frequency-domain feature group, and form a historical feature matrix with the historical operation data;

[0158] Fault analysis is performed on the feature matrix and the historical feature matrix based on the input layer, the first hidden layer, the second hidden layer, and the output layer in the preset analysis model to obtain the first analysis result. Among them, the output function of the output layer is:

[0159]

[0160] Among them, a[x] represents the output value of the x-th node in the output layer of the preset analysis model, and g k , g1, and g2 respectively represent the activation functions of the output layer, the first hidden layer, and the second hidden layer. T2 represents the number of nodes in the second hidden layer, and w xj2 represents the third weight between the second hidden layer and the output layer. T1 represents the number of nodes in the first hidden layer, and w j2j1 represents the second weight between the first hidden layer and the second hidden layer. T0 represents the number of nodes in the input layer, and w j1i represents the first weight between the input layer and the first hidden layer. p 1i represents the i-th row of the feature matrix, and p 2i represents the i-th row of the historical feature matrix. B j1 , B j2 , B x respectively represent the bias values of the first hidden layer, the second hidden layer, and the output layer.

[0161] Further, the first weight, the second weight, and the third weight are generated through a preset update formula group during the training process of the preset analysis model. The preset update formula group is:

[0162]

[0163] Among them, w’ j1i represents the updated first weight, w’ j2j1 represents the updated second weight, w’ xj2 represents the updated third weight. μ represents the learning rate during the training process of the preset analysis model. p 3i represents the i-th row of the training matrix during the training process of the preset analysis model. δ yj1 is the training error rate on the first hidden layer during the y-th training of the preset analysis model. δ yj2 is the training error rate on the second hidden layer during the y-th training of the preset analysis model. δ yx is the training error rate on the output layer during the y-th training of the preset analysis model.

[0164] Further, the on-vehicle PHM module includes sub-modules corresponding to each carriage of the train respectively. Each of the sub-modules transmits the operation data of each carriage collected based on Ethernet networking to the nodes of the Ethernet train backbone network, and is transmitted by the Ethernet train backbone network to the on-vehicle controller;

[0165] The step of performing fault monitoring on the train according to the first analysis result includes:

[0166] According to the node identifier carried in each vibration signal, search for the target carriage corresponding to the first analysis result;

[0167] According to the target carriage, the fault type in the first analysis result, and the type parameter corresponding to the fault type, generate monitoring information for performing fault monitoring on the motor shaft of the running gear in the target carriage, and output the monitoring information.

[0168] Further, the cold operation data includes the image detection data and the sensing detection data of the wheel treads in the train. The step of analyzing the cold operation data and the train parameters to generate a second analysis result for performing fault prediction on the train includes:

[0169] Perform smoothing filtering and edge extraction processing on each depth image in the image detection data to obtain a plurality of tread edge images, and convert each of the tread edge images into a binary edge image;

[0170] According to the coordinate values of the pixels in each binary edge image in the preset coordinate axis, generate a pixel matrix corresponding to each binary edge image, and perform a difference operation between each pixel matrix and the reference pixel matrix corresponding to the wheel tread to obtain a first operation matrix;

[0171] Perform a summation operation on each of the first operation matrices to obtain a second operation matrix, and generate a graphic analysis result of the wheel tread according to the second operation matrix;

[0172] Generate a sensing analysis result of the wheel tread according to the vibration voltage value, the instantaneous speed value in the sensing detection data, and the train parameters;

[0173] Generate the second analysis result to perform fault prediction on the wheel tread according to the graphic analysis result and the sensing analysis result.

[0174] Further, the step of generating a sensing analysis result of the wheel tread according to the vibration voltage value, the instantaneous speed value in the sensing detection data, and the train parameters includes:

[0175] Determine whether the vibration voltage value is greater than a preset threshold. If it is greater than the preset threshold, determine the operating section corresponding to the train tread, and search for the section coefficient corresponding to the operating section, as well as the first instrument correction coefficient and the second instrument correction coefficient corresponding to the vibration voltage value and the instantaneous speed value respectively;

[0176] Calculate the vibration voltage value, the instantaneous speed value, the section coefficient, and the wheel diameter and wheel average weight in the train parameters based on a preset calculation formula to obtain the calculation result. The preset calculation formula is:

[0177]

[0178] Wherein, H represents the calculation result, D represents the wheel diameter, U represents the vibration voltage value, L represents the section coefficient, M represents the wheel average weight, V represents the instantaneous speed value, U0 represents the first instrument correction coefficient, and V0 represents the second instrument correction coefficient;

[0179] Compare the calculation result with the corresponding relationship between the preset abrasion depth interval and the abrasion grade to determine the abrasion grade corresponding to the calculation result, and generate the sensing analysis result with the calculation result and the abrasion grade.

[0180] Further, the step of generating the second analysis result based on the graph analysis result and the sensing analysis result to perform fault prediction on the wheel tread includes:

[0181] Determine whether both the graph analysis result and the sensing analysis result carry abrasion marks. If both carry abrasion marks, obtain the first abrasion depth and the first abrasion grade in the graph analysis result, and the second abrasion depth and the second abrasion grade in the sensing analysis result;

[0182] Generate a similarity value between the first abrasion depth and the second abrasion depth, and determine whether the similarity value is less than a preset similarity threshold, and whether the first abrasion grade and the second abrasion grade are the same;

[0183] If the similarity value is less than the preset similarity threshold, and the first abrasion grade and the second abrasion grade are the same, generate an average depth value between the first abrasion depth and the second abrasion depth, and generate the second analysis result based on the average depth value and the first abrasion grade or the second abrasion grade, and perform fault prediction on the wheel tread according to the second analysis result.

[0184] Further, the step of transmitting the cold operation data to the cloud server includes:

[0185] Search for the data to be redundant in the cold running data, as well as the target redundant data corresponding to the data to be redundant, and determine whether the data to be redundant is exactly the same as the target redundant data. If they are exactly the same, remove any one of the data to be redundant and the target redundant data from the cold running data to obtain the data to be transmitted;

[0186] If the data to be redundant is not exactly the same as the target redundant data, form the data to be redundant and the target redundant data together as the data to be transmitted;

[0187] Generate an encryption key based on a preset encryption algorithm, and encrypt the data to be transmitted based on the encryption key to obtain encrypted data for transmission to the cloud server, where the preset encryption algorithm is:

[0188] Select any preset prime number from a preset prime number array, and generate a first random number and a first parameter pair;

[0189] Calculate the preset prime number, the first random number, and the first parameter pair based on a first preset formula to obtain a first calculation result, and determine whether the first calculation result is a prime number. The first preset formula is:

[0190] y = a * [e h *x 2 +b;

[0191] where y represents the first calculation result, a and b represent the first parameter pair, h represents the first random number, x represents the preset prime number, and [.] represents rounding down;

[0192] If the first calculation result is a prime number, generate a second random number and a second random number pair, and calculate the second random number, the second parameter pair, and the first calculation result based on the first preset formula to obtain a second calculation result;

[0193] Determine whether the second calculation result is a prime number. If it is a prime number, calculate the second calculation result based on a second preset formula to obtain a third calculation result;

[0194] Determine whether the third calculation result is a prime number. If it is a prime number, determine the third result as the first prime number;

[0195] Execute the step of selecting any preset prime number from the preset prime number array and generating a first random number, generate a second prime number, and generate an encryption key according to the first prime number and the second prime number.

[0196] The specific implementation manner of the train fault monitoring and prediction system based on Ethernet in the present invention is basically the same as each embodiment of the above-mentioned train fault monitoring and prediction method based on Ethernet, and will not be elaborated here.

[0197] An embodiment of the present invention also provides a storage medium. A control program is stored on the storage medium, and when the control program is executed by a processor, the steps of the above-described train fault monitoring and prediction method based on Ethernet are implemented.

[0198] The storage medium of the present invention may be a computer-readable storage medium, and its implementation manner is basically the same as that of each embodiment of the above-described train fault monitoring and prediction method based on Ethernet, and will not be described in detail herein.

[0199] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope protected by the claims of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, all fall within the protection scope of the present invention.

Claims

1. A train fault monitoring and prediction method based on Ethernet, characterized in that: The train includes an onboard controller and an onboard PHM module connected to the onboard controller via an Ethernet train backbone network. The train fault monitoring and prediction method includes: After receiving the train operation data transmitted by the on-board PHM module based on the Ethernet train backbone network, the on-board controller determines the target location of the train and requests to establish a communication connection with the edge server corresponding to the target location to transmit the train operation data to the edge server; The edge server pre-sets an identifier for data with high real-time requirements in the train, and screens the train operation data according to the pre-set identifier, divides the train operation data into hot operation data and cold operation data, and requests a cloud server that is communicatively connected to the edge server to obtain historical operation data corresponding to the hot operation data, and transmits the cold operation data to the cloud server; The edge server analyzes the thermal operation data and the historical operation data based on a preset analysis model to generate a first analysis result, and performs fault monitoring on the train according to the first analysis result; The cloud server searches for train parameters corresponding to the cold operation data of the train, analyzes the cold operation data and the train parameters, and generates a second analysis result to predict faults of the train.

2. The train fault monitoring and prediction method according to claim 1, characterized in that: The thermal operation data includes a vibration signal corresponding to a motor shaft in a running part of the train, and the edge server analyzes the thermal operation data and the historical operation data based on a preset analysis model, and the step of generating a first analysis result includes: Performing time domain analysis and frequency domain analysis on the vibration signal respectively to obtain an original time domain feature group and an original frequency domain feature group of the vibration signal; Filtering the vibration signal, and extracting a filtering time domain feature group and a filtering frequency domain feature group from the filtered vibration signal; The original time domain feature group, the original frequency domain feature group, the filtered time domain feature group and the filtered frequency domain feature group are formed into a feature matrix, and the historical operation data are formed into a historical feature matrix; Based on the input layer, the first hidden layer, the second hidden layer and the output layer in the preset analysis model, the feature matrix and the historical feature matrix are subjected to fault analysis to obtain the first analysis result, wherein the output function of the output layer is: ； Where a[x] represents the output value of the xth node in the output layer of the preset analysis model. , , Represent the activation functions of the output layer, the first hidden layer, and the second hidden layer respectively, T2 represents the number of nodes in the second hidden layer, represents the third weight between the second hidden layer and the output layer, T1 represents the number of nodes in the first hidden layer, represents the second weight between the first hidden layer and the second hidden layer, T0 represents the number of nodes in the input layer, represents the first weight between the input layer and the first hidden layer, represents the i-th row of the feature matrix, represents the i-th row of the historical feature matrix, Represent the bias values ​​of the first hidden layer, the second hidden layer and the output layer respectively.

3. The train fault monitoring and prediction method according to claim 2, characterized in that: The first weight, the second weight, and the third weight are generated by a preset update formula group during the training process of the preset analysis model, and the preset update formula group is: ； in, represents the first weight after update, represents the updated second weight, represents the updated third weight, Indicates the learning rate during the training of the preset analysis model. represents the i-th row of the training matrix during the training of the preset analysis model, is the training error rate of the preset analysis model for the yth training on the first hidden layer, is the training error rate of the preset analysis model for the yth training on the second hidden layer, The training error rate of the output layer of the preset analysis model for the yth training.

4. The train fault monitoring and prediction method according to claim 2, characterized in that: The on-board PHM module includes submodules corresponding to each carriage of the train, each of which transmits the collected operation data of each carriage to a node of the Ethernet train backbone network based on the Ethernet marshaling network, and then transmits the collected operation data to the on-board controller by the Ethernet train backbone network; The step of performing fault monitoring on the train according to the first analysis result comprises: According to the node identifier carried by each of the vibration signals, searching for a target carriage corresponding to the first analysis result; Based on the target car, the fault type in the first analysis result, and the type parameter corresponding to the fault type, monitoring information for fault monitoring of the motor shaft of the running part in the target car is generated and output.

5. The train fault monitoring and prediction method according to claim 1, characterized in that: The cold running data includes image detection data and sensor detection data of wheel treads in the train, and the step of analyzing the cold running data and the train parameters to generate a second analysis result to predict faults of the train includes: Performing smoothing filtering and edge extraction processing on each depth image in the image detection data to obtain a plurality of tread edge images, and converting each of the tread edge images into a binary edge image; Generate a pixel matrix corresponding to each of the binary edge images according to the coordinate values ​​of the pixels in the preset coordinate axis in each of the binary edge images, and perform a difference operation on each of the pixel matrices and a reference pixel matrix corresponding to the wheel tread to obtain a first operation matrix; Performing a sum operation on each of the first operation matrices to obtain a second operation matrix, and generating a graph analysis result of the wheel tread according to the second operation matrix; Generating a sensor analysis result of the wheel tread according to the vibration voltage value, the instantaneous speed value and the train parameter in the sensor detection data; The second analysis result is generated based on the graph analysis result and the sensor analysis result to predict the fault of the wheel tread.

6. The train fault monitoring and prediction method according to claim 5, characterized in that: The step of generating the sensor analysis result of the wheel tread according to the vibration voltage value, the instantaneous speed value and the train parameter in the sensor detection data comprises: Determine whether the vibration voltage value is greater than a preset threshold value, and if so, determine a running section corresponding to the wheel tread, and search for a section coefficient corresponding to the running section, as well as a first instrument correction coefficient and a second instrument correction coefficient corresponding to the vibration voltage value and the instantaneous speed value, respectively; The vibration voltage value, the instantaneous speed value, the section coefficient, and the wheel diameter and the average wheel weight in the train parameters are calculated based on a preset calculation formula to obtain a calculation result, and the preset calculation formula is: ； Among them, H represents the calculation result, D represents the wheel diameter, U represents the vibration voltage value, L represents the road section coefficient, M represents the average wheel weight, and V represents the instantaneous speed value. represents the first instrument correction factor, represents the second instrument correction factor; The calculation result is compared with a preset correspondence between the scratch depth interval and the scratch level to determine the scratch level corresponding to the calculation result, and the calculation result and the scratch level are combined to generate the sensing analysis result.

7. The train fault monitoring and prediction method according to claim 5, characterized in that: The step of generating the second analysis result to predict the wheel tread fault according to the graph analysis result and the sensor analysis result comprises: Determine whether the image analysis result and the sensor analysis result both carry a scratch mark, and if both carry a scratch mark, obtain a first scratch depth and a first scratch level in the image analysis result, and a second scratch depth and a second scratch level in the sensor analysis result; generating a similarity value between the first scratch depth and the second scratch depth, and determining whether the similarity value is less than a preset similarity threshold, and whether the first scratch level and the second scratch level are the same; If the similarity value is less than a preset similarity threshold, and the first abrasion level and the second abrasion level are the same, an average depth value between the first abrasion depth and the second abrasion depth is generated, and the second analysis result is generated based on the average depth value and the first abrasion level or the second abrasion level, and a fault prediction is performed on the wheel tread according to the second analysis result.

8. A train fault monitoring and prediction system based on Ethernet, characterized in that: The Ethernet-based train fault monitoring and prediction system includes a memory, a processor, a communication bus, and a control program stored in the memory: The communication bus is used to realize the connection and communication between the processor and the storage; The processor is used to execute the control program to implement the steps of the Ethernet-based train fault monitoring and prediction method as described in any one of claims 1-7.

9. A storage medium, characterized in that: The storage medium stores a control program, and when the control program is executed by the processor, the steps of the Ethernet-based train fault monitoring and prediction method as described in any one of claims 1-7 are implemented.

Images