A real-time detection method and device for train running gear bearing failure and a train

Abstract

The application provides a train running gear bearing fault real-time detection method and device and a train, which can be applied to the field of rail transit technology. The method comprises the following steps: acquiring a characteristic value of a real-time vibration acceleration signal of a train end running gear bearing; dynamically matching the characteristic value of the simulation vibration acceleration signal with the characteristic value of the real-time vibration acceleration signal to obtain a nonlinear relationship between the characteristic value of the real-time vibration acceleration signal and a test fault defect size; processing an estimated fault defect size obtained by inversion of the nonlinear relationship based on a fault displacement excitation, a bearing radial clearance and attribute parameters, real-time working condition parameters and a real-time load spectrum of the running gear bearing to obtain a vibration acceleration initial simulation signal, performing real-time transfer path analysis on the vibration acceleration initial simulation signal to obtain a vibration acceleration simulation signal; and performing real-time processing on the vibration acceleration simulation signal received by the train end to obtain a fault real-time detection result of the running gear bearing.

Description

Technical Field

[0001] This application relates to the field of rail transit technology, specifically to a method, device, and train for real-time detection of bearing failures in the running gear of a train. Background Technology

[0002] With the rapid development of my country's rail transit towards high speed and high density, the health of the train running gear, as the core component that supports the car body and ensures safe operation, directly affects the operational safety and efficiency of the entire line. Among these components, running gear bearings are critical and prone to failure due to the long-term exposure to complex alternating loads. Traditional maintenance models based on fixed cycles or reactive repairs are no longer sufficient to meet the stringent requirements of modern rail transit for high reliability and low maintenance costs. Therefore, the industry urgently needs to transform and upgrade towards predictive maintenance.

[0003] In realizing the concept of this application, the inventors discovered at least the following problems in the related technologies: In the existing predictive maintenance schemes for running gear bearings, the massive amounts of data from the train cannot be transmitted to the ground data center in real time, resulting in the inability to mine the massive amounts of data in a timely, effective, and sufficient manner; the existing digital twin models of running gear bearings are mostly statically constructed and cannot dynamically update parameters as defects evolve during bearing operation, leading to increased deviations between simulation and actual conditions and a disconnect between the model and the physical entity; in addition, the existing predictive maintenance schemes for running gear bearings are mostly one-way data uploads and offline analysis on the ground, and the analysis results cannot be transmitted back to the vehicle in a timely manner, making it difficult to form a closed loop of perception-decision-execution, which restricts the real-time performance and intelligence level of operation and maintenance. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus and train for real-time detection of bearing failure in the running gear of a train, to at least solve one of the existing problems.

[0005] One aspect of this application provides a real-time detection method for a train running gear bearing fault, comprising: acquiring the characteristic values ​​of a real-time vibration acceleration signal of the running gear bearing at the train end; dynamically matching the characteristic values ​​of a simulated vibration acceleration signal that dynamically changes with the size of the test fault defect with the characteristic values ​​of the real-time vibration acceleration signal to obtain a nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect; processing the estimated fault defect size obtained by inversion from the nonlinear relationship based on fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running gear bearing to obtain an initial simulated vibration acceleration signal, and performing real-time transmission path analysis on the initial simulated vibration acceleration signal to obtain a simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running gear bearing; and processing the simulated vibration acceleration signal received at the train end in real time to obtain the real-time fault detection result of the running gear bearing.

[0006] According to an embodiment of this application, the characteristic value of the real-time vibration acceleration signal of the running gear bearing received above includes: calling a vibration acceleration sensor to collect the original real-time vibration acceleration signal of the running gear bearing in real time; calling on-board fault prediction and health management equipment to preprocess the original real-time vibration acceleration signal to obtain a preprocessed real-time vibration acceleration signal; calculating the root mean square of the preprocessed real-time vibration acceleration signal to obtain the characteristic value of the real-time vibration acceleration signal, wherein the characteristic value of the real-time vibration acceleration signal is sent to the ground data center through an on-board wireless transmission device.

[0007] According to an embodiment of this application, the above-mentioned invocation of the vehicle fault prediction and health management device to preprocess the original real-time vibration acceleration signal to obtain the preprocessed real-time vibration acceleration signal includes: invoking the vehicle fault prediction and health management device to perform noise reduction and filtering processing on the original real-time vibration acceleration signal to obtain the preprocessed real-time vibration acceleration signal.

[0008] According to an embodiment of this application, the above-mentioned dynamic matching of the characteristic values ​​of the simulated vibration acceleration signal that dynamically changes with the size of the test fault defect with the characteristic values ​​of the real-time vibration signal to obtain the nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect includes: dynamically adjusting the size of the test fault defect; calling the bearing digital twin defect evolution model to generate a simulated vibration acceleration signal that dynamically adjusts with the size of the test fault defect; calculating the characteristic values ​​of the simulated vibration acceleration signal; and dynamically matching the characteristic values ​​of the simulated vibration acceleration signal with the characteristic values ​​of the real-time vibration signal to obtain the nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect.

[0009] According to an embodiment of this application, the above-mentioned dynamic adjustment of the test fault defect size includes: dynamic adjustment of the test fault defect size based on the bisection method; wherein, the bearing digital twin defect evolution model is constructed based on the error backpropagation neural network.

[0010] According to an embodiment of this application, the above-mentioned processing of the estimated fault defect size obtained by nonlinear relationship inversion based on fault displacement excitation, bearing radial clearance, attribute parameters of the running bearing, real-time operating parameters, and real-time load spectrum to obtain the initial simulated vibration acceleration signal includes: real-time initialization of the bearing digital twin mechanism model based on fault displacement excitation, bearing radial clearance, attribute parameters of the running bearing, real-time operating parameters, and real-time load spectrum to obtain the initialized bearing digital twin mechanism model; and processing the estimated fault defect size using the initialized bearing digital twin mechanism model to obtain the initial simulated vibration acceleration signal.

[0011] According to an embodiment of this application, the above-mentioned bearing digital twin mechanism model is constructed based on a two-degree-of-freedom bearing vibration differential equation.

[0012] According to an embodiment of this application, the above-mentioned analysis of the real-time transmission path of the initial simulated vibration acceleration signal to obtain a simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running gear bearing includes: performing real-time analysis of the signal transmission path between the vibration acceleration sensor and the running gear bearing to determine the path effect of real-time signal transmission, wherein the vibration acceleration sensor is arranged on the running gear bearing; and processing the real-time transmission path effect of the initial simulated vibration acceleration signal in the bearing digital twin mechanism model according to the path effect of real-time signal transmission to obtain the simulated vibration acceleration signal.

[0013] According to an embodiment of this application, the above-mentioned processing of the vibration acceleration simulation signal received by the train end to obtain the real-time fault detection result of the running gear bearing includes: the train end acquiring the vibration acceleration simulation signal sent by the ground data center in real time; calling the trained bearing digital twin diagnostic large model to process the vibration acceleration simulation signal to obtain the real-time fault detection result of the running gear bearing, wherein the trained bearing digital twin diagnostic large model is deployed on the train end.

[0014] According to an embodiment of this application, the above-mentioned process of calling the trained bearing digital twin diagnostic model to process the vibration acceleration simulation signal and obtain the real-time fault detection result of the running gear bearing includes: preprocessing the vibration acceleration simulation signal to obtain a preprocessed vibration acceleration simulation signal; using the trained bearing digital twin diagnostic model to extract features from the preprocessed vibration acceleration simulation signal to obtain vibration acceleration features; using the trained bearing digital twin diagnostic model to perform pooling processing on the vibration acceleration features to obtain pooled vibration acceleration features; using the trained bearing digital twin diagnostic model to perform linear mapping processing on the pooled vibration acceleration features to obtain mapped vibration acceleration features; and using the trained bearing digital twin diagnostic model to classify the mapped vibration acceleration features to obtain the real-time fault detection result of the running gear bearing.

[0015] According to an embodiment of this application, the bearing digital twin diagnostic model that has been trained is obtained through the following operations: constructing a hybrid training dataset using the actual vibration acceleration signal and the simulated vibration acceleration signal of the running bearing; training the bearing digital twin diagnostic model using the hybrid training dataset to obtain the initially trained bearing digital twin diagnostic model; and performing low-rank adaptive parameter fine-tuning on the initially trained bearing digital twin diagnostic model to obtain the fully trained bearing digital twin diagnostic model.

[0016] Another aspect of this application provides a real-time detection device for train running gear bearing faults, comprising: an eigenvalue acquisition module for acquiring eigenvalues ​​of real-time vibration acceleration signals of the train-end running gear bearing; a nonlinear relationship acquisition module for dynamically matching the eigenvalues ​​of simulated vibration acceleration signals that dynamically change with the size of the test fault defect with the eigenvalues ​​of the real-time vibration acceleration signals to obtain a nonlinear relationship between the eigenvalues ​​of the real-time vibration acceleration signals and the size of the test fault defect; an analog signal acquisition module for processing the estimated fault defect size obtained by inversion of the nonlinear relationship based on fault displacement excitation, bearing radial clearance, and attribute parameters, real-time operating parameters, and real-time load spectrum of the running gear bearing to obtain an initial simulated vibration acceleration signal, and performing real-time transmission path analysis on the initial simulated vibration acceleration signal to obtain a simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running gear bearing; and a real-time fault detection module for real-time processing of the simulated vibration acceleration signal received at the train end to obtain the real-time fault detection result of the running gear bearing.

[0017] Another aspect of this application provides a train including a real-time detection device for bearing failure in the train running gear as described above.

[0018] Another aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the method as described above.

[0019] Another aspect of this application provides a computer-readable storage medium storing computer-executable instructions that, when executed, are used to implement the method described above.

[0020] Another aspect of this application provides a computer program product comprising computer-executable instructions which, when executed, are used to implement the method described above.

[0021] According to the embodiments of this application, because the technical means of receiving vibration characteristic values ​​in real time, constructing nonlinear relationships by dynamically matching simulation and measured characteristic values, fusing multi-physical parameter inversion and combining transmission path analysis to generate high-precision analog signals, and processing signals in real time to realize fault detection are adopted, the technical problems of insufficient data value mining, disconnect between simulation process and physical entity, and insufficient closed-loop feedback and real-time performance are at least partially overcome. Thus, the technical effects of dynamic synchronization of simulation process, high simulation fidelity, continuous and sufficient data supply, real-time and accurate detection, and improved intelligent operation and maintenance level are achieved. Attached Figure Description

[0022] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0023] Figure 1 An exemplary system architecture diagram is shown for a real-time detection method for train running gear bearing failure according to an embodiment of this application.

[0024] Figure 2 A flowchart is shown for a real-time detection method for bearing failure in the running gear of a train according to an embodiment of this application.

[0025] Figure 3 A diagram illustrating the construction process of a bearing digital twin defect evolution model according to an embodiment of this application is shown.

[0026] Figure 4 A diagram illustrating the process of constructing a digital twin mechanism model of a bearing according to an embodiment of this application is shown.

[0027] Figure 5 A diagram illustrating the process of constructing a large-scale digital twin diagnostic model for bearings according to an embodiment of this application is shown.

[0028] Figure 6 A flowchart of a real-time detection method for train running gear bearing faults based on vehicle-to-ground interaction, according to another embodiment of this application, is shown.

[0029] Figure 7 A block diagram of a real-time detection device for train running gear bearing failure according to an embodiment of this application is shown.

[0030] Figure 8 A block diagram of an electronic device suitable for implementing a real-time detection method for bearing failures in the running gear of a train, according to an embodiment of this application, is shown. Detailed Implementation

[0031] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0032] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0033] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0034] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0035] Due to the rapid development of the rail transit industry, the demand for predictive maintenance of train running gear bearings is becoming increasingly strong. However, in the practice of implementing predictive maintenance of running gear bearings, existing technical solutions face the following problems: (1) Insufficient data value mining: Although modern trains are equipped with advanced sensing systems that can collect massive amounts of real-time vibration data, due to the limited bandwidth and stability of wireless transmission between the train and the ground, it is neither economical nor realistic to transmit all the original data to the ground in real time. This results in a large amount of "data rich mine" containing equipment status information being trapped on the train, and the ground system cannot obtain a continuous high-value data stream for model updates, forming "data islands". (2) Disconnection between model and physical entity: Existing bearing digital twin models are mostly static or based on the initial state. During the long-term operation of the train, the defects of the bearing will continue to evolve. If the model parameters cannot be dynamically updated accordingly, the simulation results will deviate more and more from the actual state of the bearing, eventually losing its value of accurate representation and prediction, forming a "static twin" or "rigid model". (3) Closed-loop feedback and real-time challenges: An ideal intelligent operation and maintenance system should have a closed-loop capability of "state perception - model update - simulation analysis - decision feedback". Existing technical solutions mostly rely on one-way data upload and offline analysis on the ground. The analysis results are difficult to be transmitted back to the vehicle terminal in a timely manner, and a closed-loop "perception-decision-execution" cycle cannot be formed, which restricts the real-time performance and intelligence level of operation and maintenance response.

[0036] In order to at least solve one of the existing problems, this application provides a method, device and train for real-time detection of bearing failure in the running gear of a train.

[0037] Figure 1 An exemplary system architecture diagram is shown for a real-time detection method for train running gear bearing failure according to an embodiment of this application.

[0038] It is important to note that Figure 1The examples shown are merely examples of system architectures that can be applied to the embodiments of this application, in order to help those skilled in the art understand the technical content of this application, but do not mean that the embodiments of this application cannot be used in other devices, systems, environments or scenarios.

[0039] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0040] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social media platform software, etc. (for example only).

[0041] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0042] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0043] It should be noted that the real-time detection method for train running gear bearing failure provided in this application embodiment can generally be executed by server 105. Correspondingly, the real-time detection device for train running gear bearing failure provided in this application embodiment can generally be located in server 105. The real-time detection method for train running gear bearing failure provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the real-time detection device for train running gear bearing failure provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Alternatively, the real-time detection method for train running gear bearing faults provided in this application embodiment can also be executed by the first terminal device 101, the second terminal device 102, or the third terminal device 103, or by other terminal devices different from the first terminal device 101, the second terminal device 102, or the third terminal device 103. Correspondingly, the real-time detection device for train running gear bearing faults provided in this application embodiment can also be installed in the first terminal device 101, the second terminal device 102, or the third terminal device 103, or in other terminal devices different from the first terminal device 101, the second terminal device 102, or the third terminal device 103.

[0044] For example, the vibration acceleration signal to be processed can be originally stored in any one of the first terminal device 101, the second terminal device 102, or the third terminal device 103 (e.g., the first terminal device 101, but not limited thereto), or it can be stored on an external storage device and imported into the first terminal device 101. Then, the first terminal device 101 can locally execute the real-time detection method for train running gear bearing faults provided in the embodiments of this application, or send the vibration acceleration signal to be processed to other terminal devices, servers, or server clusters, and have the other terminal devices, servers, or server clusters that receive the vibration acceleration signal to be processed execute the real-time detection method for train running gear bearing faults provided in the embodiments of this application.

[0045] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0046] Figure 2 A flowchart is shown for a real-time detection method for bearing failure in the running gear of a train according to an embodiment of this application.

[0047] like Figure 2As shown, the real-time detection method for bearing failure in the running gear of the train includes operations S210~S240.

[0048] In operation S210, the characteristic value of the real-time vibration acceleration signal of the train end running gear bearing is obtained.

[0049] The ground data center receives the characteristic values ​​of the real-time vibration acceleration signal transmitted from the train. Specifically, the characteristic values ​​of the real-time vibration acceleration signal are analyzed and processed by the fault detection module in the train's onboard PHM (Prognostics and Health Management) system during operation to extract the RMS (Root Mean Square) characteristic value of the vibration acceleration signal. Then, the RMS characteristic value is transmitted in real-time to the ground data center via the train's onboard WTD (Wireless Transmission Device) network using the GPRS (General Packet Radio Service).

[0050] In operation S220, the characteristic values ​​of the simulated vibration acceleration signal, which dynamically changes with the size of the test fault defect, are dynamically matched with the characteristic values ​​of the real-time vibration acceleration signal to obtain the nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect.

[0051] The test fault defect size corresponds one-to-one with the simulated vibration acceleration signal. Therefore, when the characteristic value of the real-time vibration acceleration signal matches the characteristic value of the simulated vibration acceleration signal, the test fault defect size can be associated with the real-time vibration acceleration signal.

[0052] In operation S230, the estimated fault defect size obtained by nonlinear relationship inversion is processed based on fault displacement excitation, bearing radial clearance and the attribute parameters of the running bearing, real-time operating parameters and real-time load spectrum to obtain the initial simulated vibration acceleration signal. The initial simulated vibration acceleration signal is then analyzed in real time to obtain the simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running bearing.

[0053] Based on fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running gear bearing, it is possible to perform real-time and dynamic simulation of the running gear bearing of the train in operation, thereby improving the simulation fidelity and making the handling of estimated fault defect size more consistent with reality.

[0054] By operating S240, the simulated vibration acceleration signal received at the train end is processed in real time to obtain the real-time fault detection results of the running gear bearing.

[0055] By receiving real-time vibration acceleration signals from the running gear bearings at the train end and receiving simulated vibration acceleration signals at the train end, a closed loop is formed for real-time detection of running gear bearing faults.

[0056] According to the embodiments of this application, because the technical means of receiving vibration characteristic values ​​in real time, constructing nonlinear relationships by dynamically matching simulation and measured characteristic values, fusing multi-physical parameter inversion and combining transmission path analysis to generate high-precision analog signals, and processing signals in real time to realize fault detection are adopted, the technical problems of insufficient data value mining, disconnect between simulation process and physical entity, and insufficient closed-loop feedback and real-time performance are at least partially overcome. Thus, the technical effects of dynamic synchronization of simulation process, high simulation fidelity, continuous and sufficient data supply, real-time and accurate detection, and improved intelligent operation and maintenance level are achieved.

[0057] According to an embodiment of this application, the characteristic value of the real-time vibration acceleration signal of the running gear bearing received above includes: calling a vibration acceleration sensor to collect the original real-time vibration acceleration signal of the running gear bearing in real time; calling on-board fault prediction and health management equipment to preprocess the original real-time vibration acceleration signal to obtain a preprocessed real-time vibration acceleration signal; calculating the root mean square of the preprocessed real-time vibration acceleration signal to obtain the characteristic value of the real-time vibration acceleration signal, wherein the characteristic value of the real-time vibration acceleration signal is sent to the ground data center through an on-board wireless transmission device.

[0058] The above embodiments of this application employ the technical means of using vehicle-mounted sensors to collect raw vibration signals in real time, preprocessing them with vehicle-mounted PHM equipment to extract RMS feature values, and then transmitting them to the ground data center via vehicle-mounted wireless devices. Therefore, they at least partially overcome the technical problems of limited vehicle-to-ground wireless transmission bandwidth, uneconomical raw data transmission, and difficulty in effectively utilizing high-value state data. This achieves the technical effects of significantly reducing the amount of transmitted data, ensuring transmission stability and economy, efficiently acquiring key state features, and providing reliable data support for dynamic updates and fault inversion of the ground model.

[0059] According to an embodiment of this application, the above-mentioned invocation of the vehicle fault prediction and health management device to preprocess the original real-time vibration acceleration signal to obtain the preprocessed real-time vibration acceleration signal includes: invoking the vehicle fault prediction and health management device to perform noise reduction and filtering processing on the original real-time vibration acceleration signal to obtain the preprocessed real-time vibration acceleration signal.

[0060] The following detailed description of the scheme for receiving the characteristic values ​​of real-time vibration acceleration signals provided in the above embodiments of this application will be further explained through specific implementation methods.

[0061] The aforementioned characteristic values ​​of the real-time vibration acceleration signal received from the bearing of the running gear at the train end are related to the real-time sensing and transmission of the on-board system, specifically involving signal acquisition and preprocessing, feature extraction and compression, and wireless data transmission.

[0062] First, in the signal acquisition and preprocessing stage: vibration acceleration sensors are installed on the key bearing seats of the train running gear to continuously acquire real-time vibration acceleration signals (or raw vibration acceleration signals) during train operation. The bearing fault monitoring module in the onboard PHM host performs noise reduction and filtering on the real-time vibration acceleration signals, removing high-frequency or low-frequency interference unrelated to bearing faults, thereby improving signal quality.

[0063] Then, in the feature extraction and compression stage: the preprocessed real-time vibration acceleration signal is sent to the feature calculation unit. This application focuses on extracting the RMS feature value of the vibration signal. The RMS value is selected as the core feature because it can stably and effectively characterize the overall level of vibration energy, has a strong correlation with macroscopic degradation states such as bearing wear and fatigue, and has a small data volume, making it very suitable for narrowband wireless transmission. This module periodically (e.g., once per second or once every 5 seconds) calculates and outputs an RMS feature value, realizing data compression from high-frequency vibration waveforms to low-frequency feature scalars.

[0064] Finally, in the wireless data transmission phase: the vehicle-mounted WTD device acts as a communication gateway, using the widely covered GPRS network to send the packaged timestamps, bearing numbers, RMS values, and other data to the ground data center server in real time or near real time in a low-power, low-cost manner.

[0065] Because it employs technologies such as vehicle-mounted vibration sensors to collect signals, vehicle-mounted PHM host for noise reduction and filtering preprocessing, RMS feature extraction for data compression, and low-power, low-cost wireless transmission to the ground via GPRS, it at least partially overcomes the technical problems of large original vibration data volume, insufficient vehicle-to-ground wireless transmission bandwidth and stability, and difficulty in efficiently transmitting high-value data. This achieves the technical effects of improving signal quality, significantly compressing data volume, adapting to narrowband transmission, stably characterizing bearing degradation, and realizing efficient and reliable vehicle-to-ground data transmission.

[0066] According to an embodiment of this application, the above-mentioned dynamic matching of the characteristic values ​​of the simulated vibration acceleration signal that dynamically changes with the size of the test fault defect with the characteristic values ​​of the real-time vibration signal to obtain the nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect includes: dynamically adjusting the size of the test fault defect; calling the bearing digital twin defect evolution model to generate a simulated vibration acceleration signal that dynamically adjusts with the size of the test fault defect; calculating the characteristic values ​​of the simulated vibration acceleration signal; and dynamically matching the characteristic values ​​of the simulated vibration acceleration signal with the characteristic values ​​of the real-time vibration signal to obtain the nonlinear relationship between the characteristic values ​​of the real-time vibration acceleration signal and the size of the test fault defect.

[0067] According to an embodiment of this application, the above-mentioned dynamic adjustment of the test fault defect size includes: dynamic adjustment of the test fault defect size based on the bisection method; wherein, the bearing digital twin defect evolution model is constructed based on the error backpropagation neural network.

[0068] The embodiments of this application employ a bisection method to dynamically adjust the size of test fault defects, construct a digital twin defect evolution model of the bearing based on an error backpropagation neural network, and dynamically generate simulated vibration acceleration signals. Furthermore, they dynamically match the feature values ​​of the simulation and real-time vibration signals to establish a nonlinear relationship. Therefore, these embodiments at least partially overcome the problems of insufficient accuracy in determining the size of defects in traditional bearing fault detection, the inability of the digital twin model to fit the actual operating conditions, and the poor adaptability between the simulated signals and the real-time signals on site. This enables accurate back-calculation of the size of bearing fault defects, ensures dynamic and high-precision updates of the digital twin model, forms a closed-loop intelligent operation and maintenance system, and effectively supports the real-time fault detection and predictive maintenance of train running gear bearings.

[0069] The following specific embodiments, in conjunction with the appendix, demonstrate this process. Figure 3 The process of obtaining the nonlinear relationship provided in the above embodiments of this application will be described in further detail.

[0070] Figure 3 A diagram illustrating the construction process of a bearing digital twin defect evolution model according to an embodiment of this application is shown.

[0071] Fault size serves as a crucial input to the bearing digital twin mechanism model, significantly impacting the simulation signal output. In practical applications, obtaining bearing fault size is nearly impossible; therefore, a method for estimating fault size is proposed. The fault size is indirectly obtained by comparing the RMS values ​​of the measured and simulated signals. Based on the estimated fault size, a backpropagation (BP) neural network is used to construct the bearing digital twin fault evolution model. The construction process is as follows: Figure 3 As shown.

[0072] First, an initial defect size for the bearing is set. A simulated vibration signal corresponding to this defect is generated using a digital twin model of the bearing, while simultaneously acquiring vibration signals under actual bearing fault conditions. Next, the RMS values ​​of the simulated vibration signal and the actual fault vibration signal are calculated separately. The initially set defect size is continuously optimized using a bisection method to achieve dynamic matching of the RMS values ​​of the two signals. After matching is complete, the estimated defect size corresponding to the actual fault vibration signal is calculated through inversion. Subsequently, a digital twin defect evolution model of the bearing is trained based on a BP neural network. Finally, the nonlinear relationship between the RMS value and the estimated defect size is obtained through this model, thereby achieving the goal of quickly correlating and analyzing bearing defect sizes through vibration signals. Although the estimated defect size here does not represent the actual defect size of the bearing, it effectively reflects the evolution law of bearing defects.

[0073] According to an embodiment of this application, the above-mentioned processing of the estimated fault defect size obtained by nonlinear relationship inversion based on fault displacement excitation, bearing radial clearance, attribute parameters of the running bearing, real-time operating parameters, and real-time load spectrum to obtain the initial simulated vibration acceleration signal includes: real-time initialization of the bearing digital twin mechanism model based on fault displacement excitation, bearing radial clearance, attribute parameters of the running bearing, real-time operating parameters, and real-time load spectrum to obtain the initialized bearing digital twin mechanism model; and processing the estimated fault defect size using the initialized bearing digital twin mechanism model to obtain the initial simulated vibration acceleration signal.

[0074] Fault displacement excitation refers to the additional periodic abnormal displacement disturbance generated by repeated impacts, compressions, and scrapes at the defective parts during the operation of the running gear bearing. It is the source excitation quantity that induces fault vibration and is used to determine the intensity of the fault vibration source and whether the fault characteristics are obvious in the simulation. The bearing radial clearance refers to the maximum movable clearance that can be generated between the inner and outer rings in the radial direction (perpendicular to the center line of the shaft) under the conditions of the bearing not being installed and without load or under actual assembly and loading. It is an inherent key structural parameter of the bearing and is used to determine the basic clearance of the bearing structure and operating conditions, and to determine the underlying logic of vibration transmission and stiffness matching. Only by combining the two with the operating conditions, loads, and bearing attribute parameters can the two-degree-of-freedom vibration differential equations and digital twin mechanism models fit the actual operating state of the train.

[0075] According to an embodiment of this application, the above-mentioned bearing digital twin mechanism model is constructed based on a two-degree-of-freedom bearing vibration differential equation.

[0076] The embodiments described above in this application employ a digital twin mechanism model constructed by integrating fault displacement excitation, bearing radial clearance, bearing attribute parameters, real-time operating parameters, and real-time load spectrum to initialize the two-degree-of-freedom bearing vibration differential equation. This model is then used to process the estimated fault defect size obtained from the inversion and generate the initial simulated vibration acceleration signal. Therefore, this approach at least partially overcomes the problems of traditional simulation signal generation ignoring multi-dimensional actual parameters, poor fit between the mechanism model and actual operating conditions, and large deviations in the simulation output after defect size inversion. This improves the operating condition adaptation accuracy of the bearing digital twin mechanism model, ensures a high degree of consistency between the initial simulated vibration acceleration signal and the actual operating state, and enhances the accuracy of bearing fault simulation and fault assessment.

[0077] According to an embodiment of this application, the above-mentioned analysis of the real-time transmission path of the initial simulated vibration acceleration signal to obtain a simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running gear bearing includes: performing real-time analysis of the signal transmission path between the vibration acceleration sensor and the running gear bearing to determine the path effect of real-time signal transmission, wherein the vibration acceleration sensor is arranged on the running gear bearing; and processing the real-time transmission path effect of the initial simulated vibration acceleration signal in the bearing digital twin mechanism model according to the path effect of real-time signal transmission to obtain the simulated vibration acceleration signal.

[0078] The path transmission effect refers to the combined physical effects of amplitude attenuation, phase shift, frequency filtering, and modal amplification / suppression that occur during the transmission of the original vibration acceleration signal generated by the running gear bearing body from the fault excitation location through the bearing structure, assembly contact surface, bearing seat, housing, mounting connectors, and finally to the external vibration acceleration sensor. In other words, the vibration detected by the vibration acceleration sensor is not equal to the actual original vibration of the running gear bearing, and the path effect is the distortion and correction relationship between the two.

[0079] The embodiments described above in this application employ a technique of real-time analysis of the signal transmission path from the bearing to the vibration sensor, accurate determination of the dynamic path effect, and correction of the initial simulation signal based on this effect combined with a digital twin mechanism model. Therefore, they at least partially overcome the problems of generating only the original simulation signal of the bearing body, ignoring structural transmission distortion, resulting in large deviations between the simulation signal and the sensor's measured signal, and inaccurate matching of fault features. This eliminates the calculation error caused by the vibration transmission link, allowing the final output simulated vibration acceleration signal to closely match the actual measured characteristics on site, thereby improving the accuracy of bearing fault simulation replication, defect inversion, and fault judgment.

[0080] The following describes specific implementation methods in conjunction with appendices. Figure 4 The process of acquiring the above-mentioned vibration acceleration simulation signal provided in this application will be described in further detail.

[0081] Figure 4 A diagram illustrating the process of constructing a digital twin mechanism model of a bearing according to an embodiment of this application is shown.

[0082] The estimated defect size obtained from nonlinear relationship inversion is used as a key input parameter to dynamically update the digital twin mechanism model corresponding to the bearing. The entire process is as follows: Figure 4 As shown.

[0083] First, the basic parameters of the running gear bearing, estimated defect size, and operating condition parameters are input, and variable initialization is completed. Next, the equivalent stiffness coefficient of the running gear bearing is calculated, and then the dynamic loads from actual operation are introduced, along with the displacement excitation caused by the fault and the radial clearance of the bearing. Based on this, a two-degree-of-freedom bearing vibration differential equation is established and calculated. This series of calculations is then iterated until a pre-set simulation time is reached. At this point, the bearing vibration acceleration signal is acquired, and the vibration transmission path effect from the bearing to the sensor is analyzed. Finally, simulation results containing vibration characteristics and other information are output. Because the bearing digital twin mechanism model parameters have been updated with RMS values ​​and factors such as load spectrum, radial clearance, fault displacement excitation, and transmission path are considered, the simulation signal can reflect the vibration characteristics of the running gear bearing under the current state and operating conditions with high fidelity. Overall, the vibration behavior of the running gear bearing under fault operation is recreated through numerical simulation, providing data support for fault feature identification.

[0084] According to an embodiment of this application, the above-mentioned processing of the vibration acceleration simulation signal received by the train end to obtain the real-time fault detection result of the running gear bearing includes: the train end acquiring the vibration acceleration simulation signal sent by the ground data center in real time; calling the trained bearing digital twin diagnostic large model to process the vibration acceleration simulation signal to obtain the real-time fault detection result of the running gear bearing, wherein the trained bearing digital twin diagnostic large model is deployed on the train end.

[0085] The above embodiments of this application receive vibration acceleration simulation signals transmitted from the ground in real time at the train end, and deploy the fully trained bearing digital twin diagnostic model on the vehicle end for local computation and analysis. This can get rid of the constraints of data remote transmission delay, realize localized and real-time judgment of bearing faults, improve the real-time performance and response efficiency of fault detection, ensure rapid and accurate online fault judgment of bearings in the train running gear, and support real-time operation and maintenance early warning at the vehicle end.

[0086] According to an embodiment of this application, the above-mentioned process of calling the trained bearing digital twin diagnostic model to process the vibration acceleration simulation signal and obtain the real-time fault detection result of the running gear bearing includes: preprocessing the vibration acceleration simulation signal to obtain a preprocessed vibration acceleration simulation signal; using the trained bearing digital twin diagnostic model to extract features from the preprocessed vibration acceleration simulation signal to obtain vibration acceleration features; using the trained bearing digital twin diagnostic model to perform pooling processing on the vibration acceleration features to obtain pooled vibration acceleration features; using the trained bearing digital twin diagnostic model to perform linear mapping processing on the pooled vibration acceleration features to obtain mapped vibration acceleration features; and using the trained bearing digital twin diagnostic model to classify the mapped vibration acceleration features to obtain the real-time fault detection result of the running gear bearing.

[0087] The embodiments described above rely on a large diagnostic model to sequentially preprocess, extract features, perform pooling compression, linear mapping, and classify analog signals. This process refines and optimizes vibration data features layer by layer, enhancing effective fault information and eliminating redundant interference. It can accurately uncover deep-seated bearing fault characteristics, improve fault classification and identification accuracy, and ensure reliable fault detection results and sufficient basis for judgment in running gear bearings.

[0088] According to an embodiment of this application, the bearing digital twin diagnostic model that has been trained is obtained through the following operations: constructing a hybrid training dataset using the actual vibration acceleration signal and the simulated vibration acceleration signal of the running bearing; training the bearing digital twin diagnostic model using the hybrid training dataset to obtain the initially trained bearing digital twin diagnostic model; and performing low-rank adaptive parameter fine-tuning on the initially trained bearing digital twin diagnostic model to obtain the fully trained bearing digital twin diagnostic model.

[0089] The following describes specific implementation methods in conjunction with appendices. Figure 5 This application provides a more detailed explanation of the intelligent diagnostic process based on a large-scale digital twin diagnostic model for bearings.

[0090] Figure 5 A diagram illustrating the process of constructing a large-scale digital twin diagnostic model for bearings according to an embodiment of this application is shown.

[0091] The intelligent diagnostic process based on a large-scale digital twin diagnostic model for bearings utilizes generated simulation signals to enable advanced analysis and feeds the results back to the vehicle, forming a closed loop. A massive amount of precisely labeled simulated vibration signals (with known corresponding defect sizes and types) constitutes a perfect training and inference dataset. These are input into the ground-deployed digital twin diagnostic model for bearings for diagnostic analysis, such as... Figure 5 As shown.

[0092] First, the acquired bearing vibration simulation signals undergo data preprocessing, including data cleaning and normalization, to construct a usable dataset. Next, the large model is built, where the dataset is format-converted using convolutional, pooling, and linear layers. The results are then input into a large Transformer architecture. To enable the model to perform tasks such as fault detection, the output of the last layer of the Transformer structure is extracted and transformed to the required output dimension using a linear layer. The model is then fine-tuned. The original parameters of the large model are frozen, a Low-Rank Adaptation (LORA) module is constructed and injected into the large model, and some training parameters are unfrozen for targeted fine-tuning. Finally, the model is trained and used for diagnostic analysis, where the optimized model is used to identify and diagnose bearing faults.

[0093] The following describes another specific implementation method in conjunction with the appendix. Figure 6 This application provides a more detailed description of the real-time detection method for train running gear bearing failures.

[0094] Figure 6 A flowchart of a real-time detection method for train running gear bearing faults based on vehicle-to-ground interaction, according to another embodiment of this application, is shown.

[0095] This application provides a real-time detection method for train running gear bearing faults based on vehicle-to-ground interaction. The method deploys a bearing fault monitoring module in the onboard PHM host to process and analyze bearing vibration signals collected during operation, extracting RMS feature values ​​from the vibration signals. Subsequently, the RMS feature values ​​are transmitted in real-time to the ground system via a GPRS network using an onboard WTD device. Based on the received RMS feature values ​​and a pre-constructed digital twin defect evolution model of the running gear bearing, the ground system deduces the corresponding bearing fault defect size. This defect size serves as a key parameter for dynamically updating the digital twin mechanism model of the running gear bearing. The updated mechanism model can generate high-precision simulation signals based on the actual operating conditions of the train. These simulation signals are used for real-time characterization of bearing dynamic performance and as input for diagnostic analysis of a large diagnostic model. The analysis results can be transmitted back to the onboard system via the GPRS network. The method provided in this application forms a closed-loop intelligent operation and maintenance system, namely, a closed-loop system of data acquisition → feature extraction → ground calculation → model update → simulation and diagnosis → feedback transmission. This enables the digital twin model to be dynamically updated and maintain high accuracy, providing core technical support for predictive maintenance of train running gear bearings. Figure 6As shown, the method provided in this application mainly includes a running gear bearing vibration acceleration and train speed data acquisition module, a PHM condition monitoring module, an onboard WTD data transmission module, a bearing digital twin defect evolution model module, a bearing digital twin mechanism model module, and a bearing digital twin diagnostic large model module. Its implementation process can be decomposed into three main stages: real-time perception and transmission of the onboard system, dynamic updating and high-fidelity simulation of the ground system model, and intelligent diagnosis.

[0096] This application utilizes vehicle-to-ground interaction to feed back real-time monitored vibration characteristics (RMS) to a ground-based data center. This data is then combined with a pre-built defect evolution model to deduce the actual defect size. This key parameter is used to dynamically update the mechanistic model, enabling qualitative and quantitative assessment of faults. This allows the digital twin to remain synchronized with the current health status of the physical bearing, significantly improving the simulation fidelity and reliability of the model throughout the bearing's lifecycle. Simultaneously, the bearing digital twin model is dynamically updated using real data, and the updated high-fidelity model is used to generate simulation data, empowering the large-scale diagnostic model and forming a self-reinforcing intelligent closed loop.

[0097] The dynamically updated high-fidelity mechanism model in this application can serve as a "simulation data factory," generating high-precision simulated vibration signals under various health states and fault modes based on the actual operating conditions of the train. This provides a continuous, accurately labeled, and low-cost data source for training and validating large-scale diagnostic models, breaking through the data bottleneck in the application of artificial intelligence in the industrial field.

[0098] This application establishes a complete closed loop of "sensing-transmission-update-simulation-diagnosis-feedback". The updated model generates high-precision simulation signals, which can be used for ground depth analysis, and the diagnostic results can be transmitted back to the onboard system via GPRS. This enables the train to obtain the latest health status assessments and decision recommendations in a timely manner, greatly enhancing the real-time response capability and intelligence level of the operation and maintenance system.

[0099] This application does not involve periodic or manual model updates. Instead, updates are automatically triggered via a wireless network based on real-time monitoring of the RMS (Real Values) that characterize state degradation from the vehicle-mounted device. The updates are not based on the original signal or simple thresholds, but on defect sizes with clear physical meaning derived from a defect evolution model. This ensures the scientific rigor and high accuracy of the updates. This is the core difference between this and static models or simple parameter adjustments.

[0100] Figure 7 A block diagram of a real-time detection device for train running gear bearing failure according to an embodiment of this application is shown.

[0101] like Figure 7As shown, the real-time detection device 700 for bearing faults in the train running gear includes a feature value acquisition module 710, a nonlinear relationship acquisition module 720, an analog signal acquisition module 730, and a real-time fault detection module 740.

[0102] The feature value acquisition module 710 is used to acquire the feature values ​​of the real-time vibration acceleration signal of the train end running gear bearing.

[0103] The nonlinear relationship acquisition module 720 is used to dynamically match the feature values ​​of the simulated vibration acceleration signal that dynamically changes with the size of the test fault defect with the feature values ​​of the real-time vibration acceleration signal to obtain the nonlinear relationship between the feature values ​​of the real-time vibration acceleration signal and the size of the test fault defect.

[0104] The analog signal acquisition module 730 is used to process the estimated fault defect size obtained by nonlinear relationship inversion based on fault displacement excitation, bearing radial clearance and attribute parameters of the running bearing, real-time operating parameters and real-time load spectrum, to obtain the initial analog signal of vibration acceleration, and to perform real-time transmission path analysis on the initial analog signal of vibration acceleration to obtain the analog signal of vibration acceleration characterizing the real-time vibration characteristics of the running bearing.

[0105] The real-time fault detection module 740 is used to process the simulated vibration acceleration signal received by the train in real time to obtain the real-time fault detection results of the running gear bearing.

[0106] Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be at least partially implemented as hardware circuits, such as field-programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems-on-a-chip, systems-on-a-substrate, systems-on-package, application-specific integrated circuits (ASICs), or implemented by hardware or firmware in any other reasonable manner by integrating or packaging circuits, or implemented in any one of software, hardware, and firmware, or in a suitable combination of any of these. Alternatively, one or more of the modules, submodules, units, and subunits according to the embodiments of this application can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0107] For example, any plurality of the eigenvalue acquisition module 710, nonlinear relationship acquisition module 720, analog signal acquisition module 730, and real-time fault detection module 740 can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this application, at least one of the eigenvalue acquisition module 710, nonlinear relationship acquisition module 720, analog signal acquisition module 730, and real-time fault detection module 740 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the eigenvalue acquisition module 710, the nonlinear relationship acquisition module 720, the analog signal acquisition module 730, and the real-time fault detection module 740 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0108] It should be noted that the real-time detection device for train running gear bearing failure in the embodiments of this application corresponds to the real-time detection method for train running gear bearing failure in the embodiments of this application. For a detailed description of the real-time detection device for train running gear bearing failure, please refer to the real-time detection method for train running gear bearing failure, which will not be repeated here.

[0109] Another aspect of this application provides a train including the aforementioned real-time detection device for bearing failure in the train's running gear.

[0110] Figure 8 A block diagram of an electronic device suitable for implementing a real-time detection method for bearing failures in the running gear of a train, according to an embodiment of this application, is shown.

[0111] Figure 8 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0112] like Figure 8As shown, an electronic device 800 according to an embodiment of this application includes a processor 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage portion 808 into a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0113] RAM 803 stores various programs and data required for the operation of electronic device 800. Processor 801, ROM 802, and RAM 803 are interconnected via bus 804. Processor 801 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 802 and / or RAM 803. It should be noted that the programs may also be stored in one or more memories other than ROM 802 and RAM 803. Processor 801 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0114] According to embodiments of this application, the electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to a bus 804. The electronic device 800 may also include one or more of the following components connected to the input / output (I / O) interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the input / output (I / O) interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 810 as needed so that computer programs read from it can be installed into the storage section 808 as needed.

[0115] According to embodiments of this application, the method flow according to embodiments of this application can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 809, and / or installed from removable medium 811. When the computer program is executed by processor 801, it performs the functions defined in the system of embodiments of this application. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0116] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0117] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0118] For example, according to embodiments of this application, a computer-readable storage medium may include the ROM 802 and / or RAM 803 described above and / or one or more memories other than ROM 802 and RAM 803.

[0119] Embodiments of this application also include a computer program product, which includes a computer program containing program code for executing the methods provided in the embodiments of this application. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the real-time detection method for train running gear bearing faults provided in the embodiments of this application.

[0120] When the computer program is executed by the processor 801, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0121] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 809, and / or installed from a removable medium 811. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0122] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0123] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations are not explicitly described in this application. In particular, without departing from the spirit and teachings of this application, the features described in the various embodiments of this application can be combined and / or combined in various ways. All such combinations and / or combinations fall within the scope of this application.

[0124] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Without departing from the scope of this application, those skilled in the art can make various substitutions and modifications, all of which should fall within the scope of this application.

Claims

1. A method for real-time detection of train running gear bearing failure, characterized in that, The method includes: The characteristic values ​​of the real-time vibration acceleration signal of the bearing of the train end running gear are obtained; The feature values ​​of the simulated vibration acceleration signal, which dynamically changes with the size of the test fault defect, are dynamically matched with the feature values ​​of the real-time vibration acceleration signal to obtain the nonlinear relationship between the feature values ​​of the real-time vibration acceleration signal and the size of the test fault defect. Based on the fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running gear bearing, the estimated fault defect size obtained by the inversion of the nonlinear relationship is processed to obtain the initial simulated vibration acceleration signal. The initial simulated vibration acceleration signal is then analyzed in real time to obtain the simulated vibration acceleration signal characterizing the real-time vibration characteristics of the running gear bearing. The simulated vibration acceleration signal received at the train end is processed in real time to obtain the real-time fault detection result of the running gear bearing.

2. The method of claim 1, wherein, The characteristic values ​​of the real-time vibration acceleration signal received from the bearing of the train-end running gear include: The original real-time vibration acceleration signal of the running gear bearing is collected in real time by calling the vibration acceleration sensor; The original real-time vibration acceleration signal is preprocessed by calling the vehicle fault prediction and health management equipment to obtain the preprocessed real-time vibration acceleration signal. The root mean square of the preprocessed real-time vibration acceleration signal is calculated to obtain the characteristic value of the real-time vibration acceleration signal, wherein the characteristic value of the real-time vibration acceleration signal is transmitted to the ground data center through the vehicle-mounted wireless transmission device.

3. The method of claim 2, wherein, The original real-time vibration acceleration signal is preprocessed by the on-board fault prediction and health management equipment to obtain the preprocessed real-time vibration acceleration signal, which includes: The on-board fault prediction and health management device is invoked to perform noise reduction and filtering on the original real-time vibration acceleration signal to obtain the preprocessed real-time vibration acceleration signal.

4. The method of claim 1, wherein, The feature values ​​of the simulated vibration acceleration signal, which dynamically changes with the size of the test fault defect, are dynamically matched with the feature values ​​of the real-time vibration signal to obtain the nonlinear relationship between the feature values ​​of the real-time vibration acceleration signal and the size of the test fault defect, including: The size of the test fault defect is dynamically adjusted; The bearing digital twin defect evolution model is invoked to generate a simulated vibration acceleration signal that dynamically adjusts with the size of the test fault defect; Calculate the characteristic values ​​of the simulated vibration acceleration signal; The eigenvalues ​​of the simulated vibration acceleration signal are dynamically matched with the eigenvalues ​​of the real-time vibration signal to obtain the nonlinear relationship between the eigenvalues ​​of the real-time vibration acceleration signal and the size of the test fault defect.

5. The method of claim 4, wherein, Dynamically adjusting the size of the test fault defect includes: The size of the test fault defect is dynamically adjusted based on the dichotomy method; The bearing digital twin defect evolution model is constructed based on an error backpropagation neural network.

6. The method of claim 1, wherein, Based on the fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running gear bearing, the estimated fault defect size obtained from the inversion of the nonlinear relationship is processed to obtain the initial simulated vibration acceleration signal, including: The bearing digital twin mechanism model is initialized in real time based on the fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running part bearing to obtain the initialized bearing digital twin mechanism model. The estimated fault defect size is processed using the initialized bearing digital twin mechanism model to obtain the initial simulated vibration acceleration signal.

7. The method according to claim 6, characterized in that, The bearing digital twin mechanism model is constructed based on the two-degree-of-freedom bearing vibration differential equation.

8. The method according to claim 6, characterized in that, Real-time transmission path analysis of the initial simulated vibration acceleration signal yields a simulated vibration acceleration signal characterizing the real-time vibration features of the running gear bearing, including: The signal transmission path between the vibration acceleration sensor and the running part bearing is analyzed in real time to determine the path effect of real-time signal transmission, wherein the vibration acceleration sensor is arranged on the running part bearing; Based on the path effect of real-time signal transmission, the path effect of real-time signal transmission in the initial simulated vibration acceleration signal of the bearing digital twin mechanism model is called for processing to obtain the simulated vibration acceleration signal.

9. The method according to claim 1, characterized in that, The simulated vibration acceleration signal received at the train end is processed to obtain the real-time fault detection results of the running gear bearing, including: The train receives simulated vibration acceleration signals sent from the ground data center in real time. The trained bearing digital twin diagnostic model is invoked to process the vibration acceleration simulation signal to obtain the real-time fault detection result of the running gear bearing. The trained bearing digital twin diagnostic model is deployed at the train end.

10. The method according to claim 9, characterized in that, The trained digital twin diagnostic model for the bearing is used to process the simulated vibration acceleration signal to obtain real-time fault detection results for the running gear bearing, including: The vibration acceleration simulation signal is preprocessed to obtain the preprocessed vibration acceleration simulation signal; The trained bearing digital twin diagnostic model is invoked to extract features from the preprocessed vibration acceleration simulation signal to obtain vibration acceleration features; The trained bearing digital twin diagnostic model is invoked to perform pooling processing on the vibration acceleration features to obtain the pooled vibration acceleration features. The trained bearing digital twin diagnostic model is invoked to perform linear mapping on the pooled vibration acceleration features to obtain the mapped vibration acceleration features. The trained digital twin diagnostic model for bearings is invoked to classify the mapped vibration acceleration features, thereby obtaining the real-time fault detection results of the running gear bearing.

11. The method according to claim 9, characterized in that, The trained bearing digital twin diagnostic model was obtained through the following operations: A hybrid training dataset is constructed using the actual vibration acceleration signal of the traveling bearing and the simulated vibration acceleration signal. The bearing digital twin diagnostic model was trained using the hybrid training dataset to obtain the initially trained bearing digital twin diagnostic model. The initially trained bearing digital twin diagnostic model is fine-tuned based on low-rank adaptive parameters to obtain the trained bearing digital twin diagnostic model.

12. A real-time detection device for bearing failure in a train running gear, characterized in that, The device includes: The feature value acquisition module is used to acquire the feature values ​​of the real-time vibration acceleration signal of the bearing of the train end running gear; The nonlinear relationship acquisition module is used to dynamically match the feature values ​​of the simulated vibration acceleration signal that dynamically changes with the size of the test fault defect with the feature values ​​of the real-time vibration acceleration signal to obtain the nonlinear relationship between the feature values ​​of the real-time vibration acceleration signal and the size of the test fault defect. The analog signal acquisition module is used to process the estimated fault defect size obtained by the inversion of the nonlinear relationship based on the fault displacement excitation, bearing radial clearance, and the attribute parameters, real-time operating parameters, and real-time load spectrum of the running part bearing to obtain the initial analog signal of vibration acceleration, and to perform real-time transmission path analysis on the initial analog signal of vibration acceleration to obtain the analog signal of vibration acceleration characterizing the real-time vibration characteristics of the running part bearing. The real-time fault detection module is used to process the vibration acceleration simulation signal received by the train end in real time to obtain the real-time fault detection result of the running gear bearing.

13. A train, comprising a real-time detection device for bearing failure of the train running gear as described in claim 12.

Images