Fault detection method, system, storage medium and electronic device for c80 vehicle

Abstract

The application discloses a fault detection method of C80 vehicle, which comprises the following four steps: defining the direction of the vehicle and the end position, side position, axle position, wheel position and beam position of the vehicle; installing an electronic tag in the TFDS on one end position of the vehicle; and processing and judging the vehicle image obtained by the TFDS through a deep learning algorithm. The application also discloses a detection system for executing the above fault detection method. The application also comprises a storage medium and an electronic device. After obtaining the vehicle information, the application applies the training and reasoning of the deep learning algorithm to automatically identify the vehicle parts, automatically judge the direction of the vehicle and adjust the image direction, automatically judge the fault position of the vehicle and extract the image of the fault position, thereby improving the accuracy of fault identification and the precision of fault detection in the detection work, avoiding the missed report and false report, reducing the pressure of the vehicle inspector and improving the efficiency of the vehicle inspector.

Description

Technical Field

[0001] This invention relates to the field of railway vehicle fault detection technology, and in particular to a fault detection method, system, storage medium, and electronic device for C80 vehicles. Background Technology

[0002] The C80 type coal mine open wagon is a special open wagon designed and manufactured in my country specifically for the Datong-Qinhuangdao Railway. Its main function is to transport coal. It is also a heavy-duty vehicle designed to increase the carrying capacity of the vehicle.

[0003] TFDS (Dynamic Image Detection System for Vehicle Operation Faults) is an intelligent system integrating high-speed digital image acquisition, real-time processing of large-capacity image data, and precise positioning pattern recognition technology. It facilitates the transition of vehicle operation technical inspection from a "human-inspected and human-repaired" approach to a "machine-inspected and human-repaired" approach, and the transformation of train quality control from "human-controlled" to "machine-controlled," reducing the workload of inspectors and creating conditions for improving transportation efficiency and ensuring transportation safety. The system uses high-speed cameras to acquire images of running trains, analyzes and processes them through a computer, calculates the train's speed, identifies the train type and model, and stores images of key vehicle components needed by the system. These images are displayed on a computer window in a one-car-one-file format and can be printed and transmitted as required. Through a human-machine collaborative approach, it identifies faults such as defects, breaks, and missing parts in components and accessories such as bogies, braking systems, and couplers, thereby achieving the goal of dynamically detecting vehicle quality. The system consists of two parts: a trackside image acquisition station and a train inspection and testing center. The trackside image acquisition station consists of wheel sensors (magnetic steel assemblies), a vehicle number acquisition system, a front-end information processing unit, an image server, high-speed image acquisition equipment, light source compensation equipment, trackside equipment protection devices, and network transmission equipment. The train inspection and testing center consists of network transmission equipment, a network server, and image browsing terminals.

[0004] In practical use, TFDS is equipped with multiple detection stations along the vehicle's travel path. Each station takes pictures of the vehicle to acquire images for fault detection. Existing TFDS systems employ image comparison algorithms for processing fault images during fault detection.

[0005] In actual operation, C80 rolling stock is prone to reconfiguration and changes in direction due to complex railway conditions and varying operational requirements across lines. These changes in direction and location of faults require careful management.

[0006] The existing TFDS has problems in fault detection for C80 vehicles, which are manifested in the following ways:

[0007] 1. When the vehicle's direction changes, the fault location often changes accordingly, and the fault images obtained by the detection station are prone to differences, leading to inaccurate fault location and identification.

[0008] 2. In practical applications, due to the complexity of vehicle malfunctions, unstable vehicle direction, and large differences in malfunction images, problems such as missed malfunctions and false alarms are prone to occur during the identification of vehicle malfunctions.

[0009] 3. The TFDS centralized operation platform has drawbacks such as low efficiency, high workload and low inspection accuracy of dynamic inspection personnel, and the decision-making process depends on personal experience and the inspection results are related to the professional quality and enthusiasm of the dynamic inspection personnel. These uncertainties make it impossible to guarantee the accuracy of the operation based on the drawings. Summary of the Invention

[0010] One objective of this invention is to provide a fault detection method for a C80 vehicle, comprising the following steps:

[0011] Define the vehicle's direction and its end position, side position, axle position, wheel position, and beam position;

[0012] Using TFDS vision and electronic tags at the vehicle's location as references, the predicted bearing location of the C80 vehicle is defined in TFDS, with the electronic tag installed at one end of the vehicle.

[0013] The vehicle images acquired by TFDS are processed and judged using deep learning algorithms.

[0014] Furthermore, in some preferred embodiments, defining the vehicle direction and the vehicle's end position, lateral position, axle position, wheel position, and beam position includes the following sub-steps:

[0015] Define the vehicle's longitudinal and lateral directions;

[0016] Define and differentiate the positions of the vehicle ends;

[0017] Define and differentiate the lateral positions of the vehicle; define and number the axle positions with reference to the end positions of the vehicle.

[0018] The wheel positions are defined and numbered with reference to the end and side positions;

[0019] The vehicle beams are defined and numbered based on whether they are laterally continuous, using the end position of the vehicle as a reference.

[0020] Furthermore, in some preferred embodiments, defining the vehicle direction and the vehicle's end position, lateral position, axle position, wheel position, and beam position includes the following sub-steps:

[0021] The longitudinal and lateral directions of a vehicle are defined as follows: when a vehicle is on a straight road, the longitudinal direction is defined as the connecting traction direction between the front and rear of the vehicle, and the horizontal direction perpendicular to the longitudinal direction is defined as the lateral direction.

[0022] The vehicle end positions are defined as follows: the two ends along the longitudinal direction of the vehicle are defined as parking space ends, one of which is the first parking space end and the other is defined as the second parking space end.

[0023] The lateral position of the vehicle is defined as follows: the two sides along the longitudinal direction of the vehicle are defined as the parking space sides, with one side being the first parking space side and the other side being the second parking space side.

[0024] The axle position is defined as the position of the axle in the vehicle, and the axles are numbered sequentially from the first position end to the second position end, with each axle corresponding to its own axle position;

[0025] The wheel positions are defined, with the first parking space end, the second parking space end, the first parking space side and the second parking space side as reference systems, and each wheel is numbered sequentially, with each wheel corresponding to its own wheel position.

[0026] The positions of the vehicle beams are defined as follows: For the crossbeams, when the crossbeam is a continuous beam extending from the first parking space side to the second parking space side, the position is defined using the method described above for axle positions; when the crossbeam is a non-continuous beam, the position is defined using the method described above for wheel positions. For the longitudinal beams, when the longitudinal beam is a continuous beam extending from the first parking space end to the second parking space end, the longitudinal beams are numbered sequentially from the first parking space side to the second parking space side, with each longitudinal beam corresponding to its own beam position; when the longitudinal beam is a non-continuous beam, the position is defined using the method described above for wheel positions.

[0027] Furthermore, in some preferred embodiments, the step of defining the predicted bearing position of the C80 vehicle in TFDS, using both the TFDS visual and electronic tags as references to the vehicle's location, includes the following sub-steps:

[0028] The bearing positions on both sides of the vehicle are defined and numbered according to the vehicle's visual direction of movement in TFDS;

[0029] The nominal position of the vehicle end is defined according to the distance between the vehicle end and the electronic tag. The position of the axle is defined according to the distance between the axle and the nominal end of the vehicle. Then, the positions of the bearings on both sides of the vehicle are defined and numbered according to the axle positions.

[0030] Furthermore, in some preferred embodiments, the step of defining the predicted bearing position of the C80 vehicle in TFDS, using both the TFDS visual and electronic tags as references to the vehicle's location, includes the following sub-steps:

[0031] Based on the vehicle's visual direction of movement in TFDS, with the vehicle's forward direction as the front, the four bearings on the left are defined as left 1 to left 4 from front to back, and the four bearings on the right are defined as right 1 to right 4 from front to back.

[0032] Based on the distance between the vehicle end and the electronic tag, the end of the vehicle end that is closer to the electronic tag is defined as the nominal first position end, and the other end is defined as the nominal second position end. The nominal first position end is considered the front, and the nominal second position end is considered the back. The bearing closest to the nominal first position end is defined as bearing A1 on the right and bearing A2 on the left. The bearing on the right of the second axle is defined as bearing A3 on the right and bearing A4 on the left, and so on.

[0033] Furthermore, in some preferred embodiments, processing and judging the vehicle images acquired by TFDS using deep learning algorithms includes the following sub-steps:

[0034] Vehicle image patches are sampled and clustered using a deep learning algorithm to count the image elements appearing in the vehicle images, including lines, shapes, colors, and textures.

[0035] The acquired image elements are combined using deep learning algorithms and identified as local information of a vehicle or vehicle component.

[0036] By using deep learning algorithms to annotate vehicle images and building an analysis model based on the annotations, the model identifies the image change characteristics corresponding to vehicle components and fault locations, and excludes image change characteristics of non-vehicle components and non-fault locations. This enables reasoned judgments about vehicle fault status, vehicle location, and electronic tag position.

[0037] The second objective of this invention is to provide a fault detection system for C80 vehicles, including a TFDS and a deep learning algorithm module, wherein the deep learning algorithm module is configured in the TFDS and executes the aforementioned fault detection method.

[0038] A third objective of this invention is to provide a computer-readable storage medium storing a computer program, which, when executed by one or more processors, implements the aforementioned fault detection method.

[0039] The fourth objective of this invention is to provide an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, implements the fault detection method as described in any one of claims 1-6.

[0040] Beneficial effects: This invention utilizes TFDS to acquire information about vehicles in operation, and defines vehicle parts and operating states for subsequent detection. After acquiring vehicle information, it applies deep learning algorithms for training and inference to automatically identify vehicle parts, automatically determine vehicle direction and adjust image orientation, and automatically identify and extract images of vehicle fault locations. This improves the accuracy of fault identification and the precision of fault detection in the detection process, avoids missed and false alarms, reduces the workload of inspectors, and improves their efficiency. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0042] Figure 1 This is a schematic diagram illustrating the process of a deep learning algorithm processing vehicle bearing image information in one embodiment of the present invention.

[0043] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0045] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0046] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0047] A fault detection method for C80 vehicles includes four steps: defining the location of vehicle parts, installing electronic tags, defining the location of vehicle parts during detection, and processing images using deep learning algorithms, as detailed below.

[0048] Vehicle component position definition includes defining the vehicle's direction and the positions of its ends, sides, axles, wheels, and beams.

[0049] In the steps above regarding the definition of vehicle component positions, the following definitions are established: what constitutes vehicle direction, vehicle end, vehicle side, axle position, wheel position, and vehicle position. These definitions serve as the basis for judgment when assessing a vehicle. The criteria for these definitions are based on the vehicle's own external shape and structural characteristics.

[0050] In some preferred embodiments, the vehicle location definition step includes the following sub-steps:

[0051] Define the vehicle's longitudinal and lateral directions;

[0052] Define and differentiate the positions of the vehicle ends;

[0053] Define and differentiate the lateral positions of the vehicle; define and number the axle positions with reference to the end positions of the vehicle.

[0054] The wheel positions are defined and numbered with reference to the end and side positions;

[0055] The vehicle beams are defined and numbered based on whether they are laterally continuous, using the end position of the vehicle as a reference.

[0056] The above provides a specific definition method.

[0057] In some other preferred embodiments, the vehicle location definition step includes the following sub-steps:

[0058] The longitudinal and lateral directions of a vehicle are defined as follows: when a vehicle is on a straight road, the longitudinal direction is defined as the connecting traction direction between the front and rear of the vehicle, and the horizontal direction perpendicular to the longitudinal direction is defined as the lateral direction.

[0059] The vehicle end positions are defined as follows: the two ends along the longitudinal direction of the vehicle are defined as parking space ends, one of which is the first parking space end and the other is defined as the second parking space end.

[0060] The lateral position of the vehicle is defined as follows: the two sides along the longitudinal direction of the vehicle are defined as the parking space sides, with one side being the first parking space side and the other side being the second parking space side.

[0061] The axle position is defined as the position of the axle in the vehicle, and the axles are numbered sequentially from the first position end to the second position end, with each axle corresponding to its own axle position;

[0062] The wheel positions are defined, with the first parking space end, the second parking space end, the first parking space side and the second parking space side as reference systems, and each wheel is numbered sequentially, with each wheel corresponding to its own wheel position.

[0063] The positions of the vehicle beams are defined as follows: For the crossbeams, when the crossbeam is a continuous beam extending from the first parking space side to the second parking space side, the position is defined using the method described above for axle positions; when the crossbeam is a non-continuous beam, the position is defined using the method described above for wheel positions. For the longitudinal beams, when the longitudinal beam is a continuous beam extending from the first parking space end to the second parking space end, the longitudinal beams are numbered sequentially from the first parking space side to the second parking space side, with each longitudinal beam corresponding to its own beam position; when the longitudinal beam is a non-continuous beam, the position is defined using the method described above for wheel positions.

[0064] In the above definition of vehicle end, the end with the human-powered brake or the end in the direction of the brake cylinder piston rod extension is defined as the first vehicle end. When the above meaning is not applicable, it shall be specified by the design department or based on the markings painted on the vehicle.

[0065] The above provides another specific definition method.

[0066] Install the electronic tag by installing the electronic tag from TFDS at one end of the vehicle.

[0067] In the above steps for installing electronic tags, each vehicle corresponds to one electronic tag, and the electronic tag stores information about the corresponding vehicle, including the location definitions of various parts of the vehicle body. The electronic tag transmits the vehicle information to the TFDS's computing and processing unit via radio frequency technology.

[0068] The vehicle parts are defined in the detection process, with TFDS vision and electronic tags as references respectively. The predicted bearing position of the C80 vehicle is defined separately in TFDS.

[0069] In the step of defining the vehicle's location during inspection, since TFDS acquires vehicle images and information through camera capture and electronic tags, and uses these images and information for subsequent fault detection and judgment, it is necessary to redefine the vehicle's location based on TFDS vision and the electronic tag's location. This definition exists within TFDS and is entirely the basis for inspection judgment.

[0070] In some preferred embodiments, defining the location of the vehicle component during detection includes the following sub-steps:

[0071] The bearing positions on both sides of the vehicle are defined and numbered according to the vehicle's visual direction of movement in TFDS;

[0072] The nominal position of the vehicle end is defined according to the distance between the vehicle end and the electronic tag. The position of the axle is defined according to the distance between the axle and the nominal end of the vehicle. Then, the positions of the bearings on both sides of the vehicle are defined and numbered according to the axle positions.

[0073] The above provides a specific method for defining the position of vehicle parts during detection in TFDS.

[0074] In some preferred embodiments, defining the location of the vehicle component during detection includes the following sub-steps:

[0075] Based on the vehicle's visual direction of movement in TFDS, with the vehicle's forward direction as the front, the four bearings on the left are defined as left 1 to left 4 from front to back, and the four bearings on the right are defined as right 1 to right 4 from front to back.

[0076] Based on the distance between the vehicle end and the electronic tag, the end of the vehicle end that is closer to the electronic tag is defined as the nominal first-position end, and the other end is defined as the nominal second-position end. The nominal first-position end is taken as the front, and the nominal second-position end is taken as the back. The bearing closest to the nominal first-position end is defined as bearing A1 on the right and bearing A2 on the left. The bearing on the right of the second axle is defined as bearing A3 on the right and bearing A4 on the left, and so on, until A8 is defined.

[0077] The above provides a specific method for defining the position of vehicle parts during detection in TFDS.

[0078] Deep learning algorithms are used to process images, specifically vehicle images acquired by TFDS.

[0079] Deep learning algorithms can achieve intelligent effects through training and perform inference based on the training results. Therefore, deep learning algorithms can achieve the same training effect by acquiring a large number of vehicle images.

[0080] In some preferred embodiments, deep learning algorithms process images including the following sub-steps:

[0081] Vehicle image patches are sampled and clustered using a deep learning algorithm to count the image elements appearing in the vehicle images, including lines, shapes, colors, and textures.

[0082] The acquired image elements are combined using deep learning algorithms and identified as local information of a vehicle or vehicle component.

[0083] By using deep learning algorithms to annotate vehicle images and building an analysis model based on the annotations, the model identifies the image change characteristics corresponding to vehicle components and fault locations, and excludes image change characteristics of non-vehicle components and non-fault locations. This enables reasoned judgments about vehicle fault status, vehicle location, and electronic tag position.

[0084] If a fault image identified by a detection station has a different orientation than the fault image identified by the previous detection station, the fault image will be automatically adjusted to the same orientation before the dynamic inspection operator continues to view the image, thus improving manpower efficiency.

[0085] The above describes a method that uses deep learning algorithms to cluster a large number of small image patches on a vehicle, statistically analyzing the most frequent patterns of local changes in lines, shapes, colors, and textures. In slightly larger image patches, based on combinations of these most frequent local change patterns, local information about components, such as bolts and rounded edges, is identified. In even larger image patches, guided by data annotations (such as component location and fault status), a statistical model is automatically built to identify local change patterns corresponding to components or faults. Simultaneously, local change patterns that are not related to components or faults are excluded, thus acquiring features related to the components.

[0086] like Figure 1 As shown, in the process of identifying a bearing image, the image is processed in three stages using deep learning algorithms, and is then identified as a rolling bearing.

[0087] In practical application, this invention utilizes TFDS to acquire information about vehicles in operation, while simultaneously defining vehicle parts and operating states for subsequent inspection. Electronic tags on the vehicle send vehicle information to the TFDS storage unit. After acquiring vehicle information, TFDS employs deep learning algorithms for training and inference to automatically identify vehicle parts, determine vehicle direction and adjust image orientation, and automatically identify and extract images of faulty parts. This improves the accuracy of fault identification and detection during inspection, avoids missed and false alarms, reduces the workload of inspectors, and increases their efficiency.

[0088] A fault detection system for a C80 vehicle includes a TFDS and a deep learning algorithm module, wherein the deep learning algorithm module is configured in the TFDS and executes the fault detection method described above.

[0089] A computer-readable storage medium storing a computer program that, when executed by one or more processors, implements the fault detection method.

[0090] An electronic device includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the fault detection method is implemented.

[0091] This invention applies deep learning algorithms to TFDS (Fault Detection and Disaster Management). Based on deep learning technology, it improves the accuracy and eliminates omissions in large-scale vehicle inspections. Furthermore, the images presented to inspectors are consistent and not limited by changes in vehicle direction, thus reducing manual efficiency. It can even replace manual fault identification. In practical implementation, this invention has significant practical implications for improving vehicle safety.

[0092] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method of fault detection for a C80 vehicle, characterized by It comprises the following steps: The vehicle direction and the end position, side position, axle position, wheel position and beam position of the vehicle are defined; The predicted bearing positions of the C80 vehicle are defined in the TFDS with the TFDS vision and the electronic tag at the vehicle position as the reference respectively, wherein the electronic tag is installed at one of the end positions of the vehicle; wherein the definition of the predicted bearing positions of the C80 vehicle in the TFDS with the TFDS vision and the electronic tag at the vehicle position as the reference respectively comprises the following sub-steps: the bearing positions on both sides of the vehicle are defined and numbered according to the visual running direction of the vehicle in the TFDS; the nominal position of the end of the vehicle is defined according to the distance between the end of the vehicle and the electronic tag, and the axle position is defined according to the distance between the axle and the nominal end of the vehicle, and the bearing positions on both sides of the vehicle are defined and numbered according to the axle position; The vehicle image obtained by the TFDS is processed and judged by a deep learning algorithm; wherein the vehicle image obtained by the TFDS is processed and judged by a deep learning algorithm, which comprises the following sub-steps: the vehicle image block is sampled and clustered by a deep learning algorithm, and the image elements appearing in the vehicle image are counted, and the image elements include lines, shapes, colors and textures; the obtained image elements are combined by a deep learning algorithm, and are judged as local information of the vehicle or vehicle parts; the vehicle image is data-labeled by a deep learning algorithm, and an analysis model is established according to the labeling, the image change characteristics corresponding to the vehicle parts and fault positions are found out, and the image change characteristics of non-vehicle parts and non-fault positions are excluded, thereby realizing the inferential judgment of the vehicle fault state, vehicle direction and electronic tag position; if the fault image recognized by one detection station is different from the direction of the fault image recognized by the previous detection station, the fault image is automatically adjusted to the same direction by a deep learning algorithm.

2. The method of fault detection for a C80 vehicle of claim 1, wherein, The definition of the vehicle direction and the end position, side position, axle position, wheel position and beam position of the vehicle comprises the following sub-steps: The vehicle longitudinal direction and the vehicle transverse direction are defined; The end position of the vehicle is defined and distinguished; The side position of the vehicle is defined and distinguished; the position of the axle is defined and numbered with reference to the end position of the vehicle; The wheel position is defined and numbered with reference to the end position and side position; The beam is defined and numbered with reference to the end position of the vehicle and according to whether the beam itself is side-through.

3. The method of fault detection for a C80 vehicle of claim 1, wherein, The definition of the vehicle direction and the end position, side position, axle position, wheel position and beam position of the vehicle comprises the following sub-steps: The vehicle longitudinal direction and the vehicle transverse direction are defined, when the vehicle is located on a straight line, the connection traction direction from front to back of the vehicle is defined as the vehicle longitudinal direction, and the horizontal direction perpendicular to the vehicle longitudinal direction is defined as the vehicle transverse direction; The end position of the vehicle is defined, the two ends of the vehicle along the vehicle longitudinal direction are defined as the end position, and one of them is defined as the first end position, and the other is defined as the second end position; The side position of the vehicle is defined, and the two sides of the vehicle along the longitudinal direction of the vehicle are defined as the parking space sides, one of which is the first parking space side and the other is the second parking space side; The position of the axle is defined, and the position of the axle in the vehicle is defined, and the second parking space end is sequentially numbered from the first parking space end, and each axle corresponds to its own axle position; The position of the wheel is defined, and the first parking space end, the second parking space end, the first parking space side and the second parking space side are used as the reference system, each wheel is sequentially numbered, and each wheel corresponds to its own wheel position; The position of the beam is defined, and when the beam is an integral beam that continuously penetrates the first parking space side and the second parking space side, the definition of the axle position is used, and when the beam is not an integral beam, the definition of the wheel position is used; the definition of the longitudinal beam position, when the longitudinal beam is an integral beam that continuously penetrates the first parking space end and the second parking space end, the longitudinal beam is sequentially numbered from the first parking space side to the second parking space side, and each longitudinal beam corresponds to its own beam position, and when the longitudinal beam is not an integral beam, the definition of the wheel position is used.

4. The method of fault detection for a C80 vehicle of claim 1, wherein, The bearing positions of the C80 vehicle in the TFDS are defined as follows: According to the visual running direction of the vehicle in the TFDS, the four bearings on the left side are defined as left 1 to left 4 from front to back, and the four bearings on the right side are defined as right 1 to right 4 from front to back. According to the distance between the end of the vehicle and the electronic tag, the end of the vehicle that is closer to the electronic tag is defined as the nominal one end, and the other end is defined as the nominal two end. The nominal one end is in front of the nominal two end, and the nominal one end is closest to the nominal one end, and the right side bearing of the first axle is defined as bearing A1, and the left side bearing is defined as bearing A2. The right side bearing of the second axle is defined as bearing A3, and the left side bearing is defined as bearing A4, and so on.

5. A fault detection system for a C80 vehicle, characterized by, The TFDS and the deep learning algorithm module are included, the deep learning algorithm module is arranged in the TFDS, and the fault detection method of any one of claims 1-4 is executed.

6. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by one or more processors to realize the fault detection method of any one of claims 1-4.

7. An electronic device, comprising: The memory and the processor are included, and the memory stores a computer program, and the computer program is executed by the processor to realize the fault detection method of any one of claims 1-4.

Images