Fault detection device, method and equipment for rail transit train high-voltage equipment

Abstract

The invention provides a fault detection device, method and equipment for rail transit train high-voltage equipment, and can be applied to the technical field of train detection. The fault detection device for the high-voltage equipment of the rail transit train comprises an audio acquisition module which comprises a servo unit and a plurality of acoustic sensors, the plurality of acoustic sensors are arranged on the servo unit, and the servo unit is configured to move under the control of a motion instruction, so that the plurality of acoustic sensors are distributed around the high-voltage equipment of the train to be detected; and the processing module is used for acquiring the sound signals acquired by the plurality of acoustic sensors respectively when the high-voltage equipment of the train to be detected is in a working state, and performing fault detection based on the plurality of sound signals to obtain a fault detection result.

Description

Technical Field

[0001] The present disclosure relates to the field of train detection technology, and more specifically, to a fault detection device, method, and equipment for high-voltage equipment of a rail transit train. Background Art

[0002] High-voltage equipment is typically installed on the roof of a train. Through electrical contact between the rooftop equipment and high-voltage cables, high-voltage power is transmitted to the train, driving its normal operation. However, since the rooftop high-voltage equipment is exposed to complex and changing operating environments for a long time, regular inspections are required to reduce the probability of train safety accidents caused by failures of the rooftop high-voltage equipment.

[0003] In the process of realizing the concept disclosed in the present invention, research has found that the existing technology usually uses a camera to shoot the high-voltage equipment on the roof and analyzes the image data obtained, which makes it difficult to accurately analyze and warn of faults of the high-voltage equipment. In addition, since the processing process of the image data is redundant and complex, it requires a large amount of memory resource space, resulting in a small application scope and low efficiency of fault analysis. Summary of the Invention

[0004] In view of this, the present disclosure provides a fault detection device, method and equipment for high-voltage equipment of a rail transit train.

[0005] One aspect of the present disclosure provides a fault detection device for high-voltage equipment of a rail transit train, comprising: an audio acquisition module, including a servo unit and multiple acoustic sensors, the multiple acoustic sensors are arranged in the servo unit, and the servo unit is configured to move under the control of a motion instruction so that the multiple acoustic sensors surround the high-voltage equipment distributed on the train to be detected; and a processing module, which is used to obtain sound signals collected by each of the multiple acoustic sensors when the high-voltage equipment of the train to be detected is in a working state, and perform fault detection based on the multiple sound signals to obtain a fault detection result.

[0006] According to an embodiment of the present disclosure, the processing module is used to extract audio features from multiple sound signals respectively to obtain multiple Mel-spectrograms, and use the mean aggregation mechanism and the maximum probability mechanism to make decisions based on the multiple Mel-spectrograms to obtain fault detection results.

[0007] According to an embodiment of the present disclosure, the processing module is used to downsample the sound signal to obtain a target sound signal, perform time domain segmentation on the target sound signal to obtain multiple time domain frame signals, perform Fourier transform on the multiple time domain frame signals respectively to obtain multiple frequency domain frame signals, and use a Mel filter group to map the multiple frequency domain frame signals to obtain a logarithmic Mel spectrum diagram.

[0008] According to an embodiment of the present disclosure, the processing module is used to utilize a mean aggregation mechanism to perform a global mean calculation on the fault probability distribution output from each time window of a plurality of logarithmic Mel-spectrograms, determine the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment, utilize a maximum probability mechanism to accumulate the fault categories with the highest fault probability in each time window of the plurality of logarithmic Mel-spectrograms, determine the number of occurrences of each of the multiple types of faults, and obtain a fault detection result based on the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment and the number of occurrences of each of the multiple types of faults.

[0009] According to an embodiment of the present disclosure, the first visual acquisition module is configured to be set relative to the first target carriage of the train to be detected. The first visual acquisition module is used to collect the appearance image of the train to be detected and send the appearance image to the processing module.

[0010] According to an embodiment of the present disclosure, the processing module is also used to determine the model of the train to be detected based on the appearance image, obtain the vehicle model characteristic parameters corresponding to the model of the train to be detected from the database, and generate motion instructions for the servo unit based on the vehicle model characteristic parameters.

[0011] According to an embodiment of the present disclosure, the processing module is used to perform dynamic range compression and distortion correction on the appearance image to obtain a preprocessed appearance image, use the image recognition model to identify the vehicle number area of ​​the preprocessed appearance image to obtain the vehicle number area coordinates, based on the vehicle number area coordinates, obtain the vehicle number area image from the preprocessed appearance image, and perform image recognition on the vehicle number area image to determine the model of the train to be detected.

[0012] According to an embodiment of the present disclosure, the processing module is also used to perform image recognition on the vehicle number area image, obtain the train number of the train to be detected, and write the fault detection result into the digital maintenance file corresponding to the train number.

[0013] According to an embodiment of the present disclosure, the remote monitoring platform is communicatively connected to the processing module. The remote monitoring platform is used to receive a query operation, determine the target train number indicated by the query operation, and send a query request for the target train number to the processing module; the processing module is also used to respond to the query request, determine the target digital maintenance file corresponding to the target train number, and send the target digital maintenance file to the remote monitoring platform, so that the remote monitoring platform can display the target digital maintenance file.

[0014] According to an embodiment of the present disclosure, the second visual acquisition module is configured to be set relative to the second target carriage of the train to be inspected, and the second target carriage includes a carriage equipped with high-voltage equipment; the second visual acquisition module is used to acquire an environmental image containing the high-voltage equipment and send the environmental image to the processing module.

[0015] According to an embodiment of the present disclosure, the processing module is also used to perform environmental safety detection based on the environmental image, obtain detection results, and when the detection results indicate that there are no abnormal objects around the high-voltage equipment, execute motion instructions to drive the servo unit to move and change the spatial distribution of multiple acoustic sensors.

[0016] According to an embodiment of the present disclosure, the processing module is further configured to feed back a warning signal to the remote monitoring platform and stop the execution of the motion instruction when the detection result indicates that there are abnormal objects around the high-voltage equipment.

[0017] Another aspect of the present disclosure provides a method for fault detection of high-voltage equipment in a rail transit train, comprising: obtaining sound signals collected by multiple acoustic sensors, wherein the multiple acoustic sensors are distributed around the high-voltage equipment of the train to be detected; and performing fault detection based on the multiple sound signals to obtain a fault detection result.

[0018] Another aspect of the present disclosure provides an electronic device, including: 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 implement the above method.

[0019] Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the above method when executed.

[0020] Another aspect of the present disclosure provides a computer program product, which includes computer-executable instructions. When the instructions are executed, they are used to implement the above method.

[0021] According to an embodiment of the present disclosure, a fault detection device for detecting high-voltage equipment on a train may include an audio acquisition module and a processing module. When the high-voltage equipment is located in a detection area provided with the audio acquisition module, a motion command is generated and sent to a servo unit in the audio acquisition module. The servo unit self-adjusts according to the received motion command so that multiple acoustic sensors provided on the servo unit can be evenly distributed around the high-voltage equipment, thereby achieving real-time dynamic adjustment of the relevant acquisition device before performing fault detection and analysis on the high-voltage equipment. This allows the multiple acoustic sensors to be evenly distributed around the high-voltage equipment, thereby improving the accuracy, efficiency, and scope of application of the collected sound signals, and can better meet the detection needs in various detection environments. The collected multiple sound signals are then transmitted to the processing module, which performs fault detection analysis and processing on the received multiple sound signals in a working state, thereby obtaining accurate detection results for the high-voltage equipment on the train in a non-contact detection manner, improving the detection efficiency and accuracy of the high-voltage equipment and the safety of the detection equipment and detection personnel, while reducing the dependence on the internal hardware performance of the train, reducing the memory resources consumed by fault detection, and improving the adaptability and application environment of fault detection. Furthermore, other detection processes related to fault detection can be added to the processing module, so that in addition to sound factors, other environmental factors can be combined to perform multi-dimensional detection of high-voltage equipment, thereby improving the accuracy of fault detection and the degree of optimization of fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0023] Figure 1 A schematic diagram schematically illustrates a fault detection device for high-voltage equipment of a rail transit train to which an embodiment of the present disclosure may be applied;

[0024] Figure 2 A schematic diagram schematically shows a remote monitoring platform to which the embodiments of the present disclosure can be applied;

[0025] Figure 3 A schematic diagram schematically illustrates a fault detection device for high-voltage equipment of a rail transit train to which another embodiment of the present disclosure may be applied;

[0026] Figure 4 A flowchart schematically illustrates a method for detecting a fault in a high-voltage device of a rail transit train to which an embodiment of the present disclosure may be applied;

[0027] Figure 5 A flowchart schematically illustrates a method for detecting a fault in a high-voltage device of a rail transit train to which another embodiment of the present disclosure may be applied;

[0028] Figure 6 A block diagram of an electronic device for a method for detecting a fault of high-voltage equipment of a rail transit train according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0029] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the detailed description below, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0030] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise," "include," etc. used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0031] 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 should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0032] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0033] In the embodiments of this disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of all data involved (including, but not limited to, user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and maintain the security of user personal information and network security.

[0034] In the embodiments of the present disclosure, the user's authorization or consent is obtained before obtaining or collecting the user's personal information.

[0035] As a core component of the EMU's energy supply, the rooftop high-voltage equipment's operating status is closely linked to the train's overall performance and operational safety. Through electrical contact between the rooftop high-voltage equipment and the high-voltage cables, high-voltage electrical energy is transmitted to the train, driving its normal operation. However, because the high-voltage cables of this equipment are typically exposed, they are exposed to complex and changing operating environments for long periods of time. They are susceptible to foreign object impacts and high-voltage discharges, which can easily damage the cable insulation and lead to breakdown accidents. Furthermore, harsh climatic conditions such as high temperature and high humidity can accelerate the aging of insulation materials, resulting in a decrease in insulation performance, which can cause equipment failures and even safety accidents.

[0036] However, existing fault detection for high-voltage rooftop equipment typically utilizes cameras to capture image data of the equipment during operation, and then uses this image data to diagnose problems such as large pantograph sparks, structural anomalies, tilt, and suspended foreign objects. In the process of implementing the above-mentioned inventive concept, it was discovered that fault detection in the prior art typically relies on analyzing image data from high-voltage rooftop equipment. However, due to the low accuracy of image data, it is difficult to accurately analyze and provide early warnings for high-voltage equipment faults. Furthermore, the image data processing process is redundant and complex, requiring a large amount of memory resources, resulting in a limited scope of application and low efficiency in fault analysis.

[0037] An embodiment of the present disclosure provides a fault detection device for high-voltage equipment of a rail transit train, comprising: an audio acquisition module, comprising a servo unit and multiple acoustic sensors, the multiple acoustic sensors being arranged on the servo unit, the servo unit being configured to move under the control of a motion instruction so that the multiple acoustic sensors surround the high-voltage equipment distributed on the train to be detected; and a processing module, for obtaining sound signals collected by each of the multiple acoustic sensors when the high-voltage equipment of the train to be detected is in a working state, and performing fault detection based on the multiple sound signals to obtain a fault detection result.

[0038] According to an embodiment of the present disclosure, a fault detection device for high-voltage equipment of a rail transit train may include an audio acquisition module and a processing module.

[0039] According to an embodiment of the present disclosure, the audio acquisition module may include a servo unit and multiple acoustic sensors, the multiple acoustic sensors are arranged in the servo unit, and the servo unit is configured to move under the control of motion instructions so that the multiple acoustic sensors surround the high-voltage equipment distributed on the train to be inspected.

[0040] When collecting sound information from high-voltage equipment, multiple audio collection modules can be installed at the guardrails of the maintenance depot top platform at the stops of the middle train and the rear train, for example, at the guardrails of the maintenance depot top platform at the stops of cars 03 and 06 of the train.

[0041] Motion commands can be directional commands for the servo units to turn left, right, or the like. These commands control the rotation of the servo units within each audio acquisition module, evenly distributing the sound sensors on the servo units around the high-voltage equipment, improving the accuracy and adaptability of the collected sound signals. Furthermore, after completing the current high-voltage equipment inspection, motion commands can be used to return multiple servo units to their standby positions, facilitating initialization for the next high-voltage equipment inspection.

[0042] According to an embodiment of the present disclosure, the processing module can be used to obtain sound signals collected by multiple acoustic sensors when the high-voltage equipment of the train to be detected is in working condition, and perform fault detection based on the multiple sound signals to obtain a fault detection result.

[0043] When the high-voltage equipment is in working condition, acoustic sensors evenly arranged around the high-voltage equipment are used to collect the sounds it makes during operation, and the collected sound signals are sent to a processing module located in the multimodal diagnostic host. The processing module performs feature analysis on multiple sound signals separately, so as to determine whether the high-voltage equipment at this time has problems such as large sparks on the pantograph, structural abnormalities, tilt, and suspended foreign objects.

[0044] According to an embodiment of the present disclosure, a fault detection device for detecting high-voltage equipment on a train may include an audio acquisition module and a processing module. When the high-voltage equipment is located in a detection area provided with the audio acquisition module, a motion command is generated and sent to a servo unit in the audio acquisition module. The servo unit self-adjusts according to the received motion command so that multiple acoustic sensors provided on the servo unit can be evenly distributed around the high-voltage equipment, thereby achieving real-time dynamic adjustment of the relevant acquisition device before performing fault detection and analysis on the high-voltage equipment. This allows the multiple acoustic sensors to be evenly distributed around the high-voltage equipment, thereby improving the accuracy, efficiency, and scope of application of the collected sound signals, and can better meet the detection needs in various detection environments. The collected multiple sound signals are then transmitted to the processing module, which performs fault detection analysis and processing on the received multiple sound signals in a working state, thereby obtaining accurate detection results for the high-voltage equipment on the train in a non-contact detection manner, improving the detection efficiency and accuracy of the high-voltage equipment and the safety of the detection equipment and detection personnel, while reducing the dependence on the internal hardware performance of the train, reducing the memory resources consumed by fault detection, and improving the adaptability and application environment of fault detection. Furthermore, other detection processes related to fault detection can be added to the processing module, so that in addition to sound factors, other environmental factors can be combined to perform multi-dimensional detection of high-voltage equipment, thereby improving the accuracy of fault detection and the degree of optimization of fault detection.

[0045] Figure 1 A schematic diagram of a fault detection device for high-voltage equipment of a rail transit train to which an embodiment of the present disclosure can be applied is shown schematically.

[0046] like Figure 1 As shown, the fault detection device for high-voltage equipment of a rail transit train may include an audio acquisition module 101 and a processing module 104. When the high-voltage equipment 106 of a train to be detected 105 needs to be detected for faults, the processing module 104 generates a motion instruction and sends it to the servo unit 102 in the audio acquisition module 101. The servo unit 102 drives the acoustic sensor 103 to rotate in response to the control of the motion instruction, so that the acoustic sensor 103 is evenly surrounded by the high-voltage equipment 106 of the train to be detected 105. The acoustic sensor 103 collects the sound signal emitted by the high-voltage equipment 106 during normal operation, and sends the sound signal to the processing module 104 for analysis and processing to obtain a fault detection result.

[0047] According to an embodiment of the present disclosure, the processing module is used to extract audio features from multiple sound signals respectively to obtain multiple Mel-spectrograms, and use the mean aggregation mechanism and the maximum probability mechanism to make decisions based on the multiple Mel-spectrograms to obtain fault detection results.

[0048] The sound signal collected by the acoustic sensor usually contains noise other than the fault sound to be detected, so the audio features of the sound signal can be extracted to extract the audio features related to fault detection and improve the detection accuracy.

[0049] For the multiple Mel-spectrograms obtained after extraction, a complementary decision-making mechanism composed of a mean aggregation mechanism and a maximum probability mechanism is introduced. The probability distribution of the fault audio can be calculated for the multiple Mel-spectrograms obtained above based on the mean aggregation mechanism. Then, based on the maximum probability mechanism, the score results of various faults are calculated according to the probability distribution of various fault audios obtained. Finally, based on the probability distribution of various fault audios and the score results of various faults, a comprehensive judgment decision is made to obtain the fault detection result of the high-voltage equipment.

[0050] Furthermore, audio feature extraction from sound signals and processing of mel-spectrograms using mean aggregation and maximum probability mechanisms can be achieved using the Mobilenet_v1 depthwise separable convolutional architecture. The Mobilenet_v1 depthwise separable convolutional architecture includes multiple convolutional layers, depthwise separable convolutional layers, and logistic layers. Multiple mel-spectrograms are input into the convolutional layers of the Mobilenet_v1 depthwise separable convolutional architecture to generate 1024-dimensional embedding vectors. The logistic layer, which incorporates mean aggregation and maximum probability mechanisms, outputs fault detection results. Before processing the mel-spectrograms using the Mobilenet_v1 depthwise separable convolutional architecture, the model architecture can be trained. Training audio features covering a variety of typical fault types and their associated noise are input into the Mobilenet_v1 depthwise separable convolutional architecture to be trained. Based on the training results, the loss value of the Mobilenet_v1 depthwise separable convolutional architecture to be trained is calculated, and its parameters are adjusted to obtain the trained Mobilenet_v1 depthwise separable convolutional architecture.

[0051] According to the embodiment of the present disclosure, the processing module is first used to extract audio features from each of the collected sound signals mixed with noise, and multiple mel-spectrograms are obtained, thereby filtering out the interfering sounds in the sound signals, thereby avoiding interference and reducing the detection accuracy when using the mean aggregation mechanism and the maximum probability mechanism to perform comprehensive feature analysis and decision-making processing on multiple mel-spectrograms. Then, a comprehensive mechanism obtained by the joint action of the mean aggregation mechanism and the maximum probability mechanism is used to perform a comprehensive calculation of the fault probability distribution and fault score based on the multiple mel-spectrograms to obtain the fault detection result of the high-voltage equipment, thereby combining the mean aggregation mechanism and the maximum probability mechanism, and then using the comprehensive decision-making mechanism obtained after the combination to make decisions based on the multiple mel-spectrograms. The results obtained after the two mechanisms are individually decided are subjected to multi-level and multi-dimensional feature analysis and decision-making, and the current high-voltage equipment fault detection situation is closed-loop diagnosed, thereby comprehensively analyzing and obtaining the current detection result, greatly improving the efficiency and accuracy of fault detection.

[0052] According to an embodiment of the present disclosure, the processing module is used to downsample the sound signal to obtain a target sound signal, perform time domain segmentation on the target sound signal to obtain multiple time domain frame signals, perform Fourier transform on the multiple time domain frame signals respectively to obtain multiple frequency domain frame signals, and use a Mel filter group to map the multiple frequency domain frame signals to obtain a logarithmic Mel spectrum diagram.

[0053] In order to filter out irrelevant noise in the sound signal, the sound signal can be downsampled first to obtain a target sound signal of a predetermined frequency, wherein the predetermined frequency can be a target sound signal in a mono format of 16kHz. Then, the target sound signal is subjected to time domain segmentation processing, and based on a preset time period and a preset compensation, the target sound signal is segmented into multiple time domain frame signals, wherein the preset segmentation time period can be 25ms, and the preset segmentation step can be 10ms. For example, after segmentation based on the preset time period and the preset compensation, the first time domain frame signal can be 0-25ms, the second time domain frame signal can be 10-35ms, and the third time domain frame signal can be 20-45ms. By using the preset segmentation step of 10ms, the segmentation window is moved every 10ms, thereby reducing the duration of each signal segment to be analyzed and processed, improving the processing speed, and by making the time domain frame signals of each adjacent frame overlap by a predetermined time period, the probability of missing signals and thus causing fault detection errors is increased.

[0054] Furthermore, short-time Fourier transforms are performed on the multiple time-domain frame signals to obtain multiple frequency-domain frame signals. Finally, the multiple frequency-domain frame signals are mapped using a Mel filter bank to obtain logarithmic Mel-spectrograms for a predetermined number of frequency bands, where the predetermined number of frequency bands can be 64. By performing the above processing on each sound signal, a logarithmic Mel-spectrogram corresponding to each sound signal is obtained.

[0055] According to an embodiment of the present disclosure, for a sound signal collected by an acoustic sensor and containing a variety of interference noises irrelevant to detection, the multiple sound signals are downsampled, time-domain segmented, and frequency-domain converted to obtain a plurality of frequency-domain frame signals, and then the multiple frequency-domain frame signals are mapped by using a Mel filter, and the features of the signal that still contains mixed redundant information and feature information after short-time Fourier transform processing are further extracted to obtain a logarithmic Mel spectrum corresponding to each sound signal, thereby realizing preprocessing of the sound signal before making a decision, so that the effective features in the sound signal are extracted through a variety of processing means, and a logarithmic Mel spectrum that can be directly input into a model for processing and analysis using a variety of decision-making mechanisms is obtained, thereby improving the accuracy and efficiency of fault detection.

[0056] According to an embodiment of the present disclosure, the processing module is used to utilize a mean aggregation mechanism to perform a global mean calculation on the fault probability distribution output from each time window of a plurality of logarithmic Mel-spectrograms, determine the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment, utilize a maximum probability mechanism to accumulate the fault categories with the highest fault probability in each time window of the plurality of logarithmic Mel-spectrograms, determine the number of occurrences of each of the multiple types of faults, and obtain a fault detection result based on the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment and the number of occurrences of each of the multiple types of faults.

[0057] By utilizing the mean aggregation mechanism of the improved model based on the YAMNet architecture, the fault probability of the frequency domain frame signal within each time window in each mel-spectrogram can be calculated separately. The fault probabilities corresponding to multiple time windows in the same mel-spectrogram are then averaged to obtain the global mean fault probability corresponding to the mel-spectrogram of a single sound signal. The same processing operations are performed on multiple mel-spectrograms to obtain the global mean fault probability corresponding to each mel-spectrogram. Based on these multiple global fault probability means, the probability of occurrence of various types of faults in high-voltage equipment can be determined.

[0058] For example, in the first time window of the first mel-spectrogram, the probability of a large pantograph spark is 0.23, and the probability of a hanging foreign object is 0.84. In the second time window, the probability of a large pantograph spark is 0.46, and the probability of a hanging foreign object is 0.11. Therefore, the global mean probability of a large pantograph spark in the first mel-spectrogram is 0.345, and the global mean probability of a hanging foreign object is 0.475. In the first time window of the second mel-spectrogram, the probability of a large pantograph spark is 0.56, and the probability of a hanging foreign object is 0.01. In the second time window, the probability of a large pantograph spark is 0.49, and the probability of a hanging foreign object is 0.03. Therefore, the global mean probability of a large pantograph spark in the first mel-spectrogram is 0.525, and the global mean probability of a hanging foreign object is 0.02. Based on the global mean probabilities of large pantograph sparks and hanging foreign objects obtained from the two Mel-spectrograms, and then performing mean calculation, the probability values ​​of various faults of high-voltage equipment can be determined, that is, the probability value of large pantograph sparks can be 0.435 and the probability value of hanging foreign objects can be 0.2475. According to the first predetermined probability threshold (greater than the first predetermined probability threshold, it can be confirmed that the fault has occurred) and the second predetermined probability threshold (greater than the second predetermined probability threshold and less than the first predetermined probability threshold, it can be confirmed that the fault may have occurred, and less than the second predetermined probability threshold, it can be confirmed that the fault has not occurred).

[0059] After calculating and processing multiple Mel-spectrograms based on the mean aggregation mechanism, the maximum probability mechanism is used to perform decision processing on the multiple Mel-spectrograms on the basis of the mean calculation using the mean aggregation mechanism to confirm the number of times each fault occurs in each frequency domain frame signal. The global fault probability mean calculated based on the mean aggregation mechanism and the number of times each fault occurs are comprehensively decided according to different weight distribution strategies to obtain the fault detection result.

[0060] According to the embodiments of the present disclosure, after pre-processing the sound signal, multiple Mel-spectrograms are obtained, and the multiple Mel-spectrograms are input into the processing module. The mean aggregation mechanism is used to perform global mean calculation on the fault probability distribution output from each time window in each Mel-spectrogram. At the same time, the maximum probability mechanism is used to calculate the number of occurrences of each type of fault based on the fault probability distribution output from each time window in each Mel-spectrogram. The mean aggregation mechanism and the maximum probability mechanism are combined to jointly determine the obtained probability value and the number of occurrences, and a comprehensive multi-dimensional decision is made on the fault of the high-voltage equipment, and then the current fault detection result is determined, thereby realizing high-precision fault detection of the high-voltage equipment, and analyzing the possible fault types and fault probabilities of the current high-voltage equipment from the collected sound signals from multiple aspects and angles.

[0061] According to an embodiment of the present disclosure, the fault detection device may further include a first visual acquisition module and a second visual acquisition module.

[0062] According to an embodiment of the present disclosure, the first visual acquisition module is configured to be set relative to the first target carriage of the train to be detected. The first visual acquisition module is used to collect the appearance image of the train to be detected and send the appearance image to the processing module.

[0063] The first visual acquisition module can be set at the guardrail of the maintenance depot top platform at the stop of the head train, for example, at the guardrail of the maintenance depot top platform at the stop of the 01 car of the train.

[0064] According to an embodiment of the present disclosure, by setting a first visual acquisition module in the corresponding detection area of ​​the train, the first visual acquisition module can be used to photograph the appearance of the train with high-voltage equipment to be inspected, and the photographed appearance image can be transmitted to the processing module, so that the processing module can first perform visual inspection based on the appearance image before performing acoustic inspection on the high-voltage equipment, and adaptively adjust the acoustic inspection based on the visual inspection feedback information of the visual inspection, thereby realizing the establishment of a high-voltage equipment fault detection process in which vision and acoustics are mutually correlated. Based on the correlation between each other, the acoustic detection is adjusted using visual feedback information, so that the acoustic sensor can collect accurate sound signals, and perform high-precision and high-efficiency closed-loop inspection on the high-voltage equipment under the condition of non-contact fault detection.

[0065] According to an embodiment of the present disclosure, the processing module can also be used to determine the model of the train to be detected based on the appearance image, obtain the vehicle model characteristic parameters corresponding to the model of the train to be detected from the database, and generate motion instructions for the servo unit based on the vehicle model characteristic parameters.

[0066] After using the first visual acquisition module to take a photo of the appearance of the train to be inspected, the photographed appearance image is transmitted to the processing module. The processing module can extract the vehicle identification area information included in the appearance image, thereby determining the model and other attribute information of the train to be inspected.

[0067] According to an embodiment of the present disclosure, the model of the train to be detected can be determined by using a processing module through the following operations: performing dynamic range compression and distortion correction on the appearance image to obtain a preprocessed appearance image, using an image recognition model to perform vehicle number area recognition on the preprocessed appearance image to obtain vehicle number area coordinates, based on the vehicle number area coordinates, intercepting a vehicle number area image from the preprocessed appearance image, and performing image recognition on the vehicle number area image to determine the model of the train to be detected.

[0068] The image recognition model can locate the area in the image that can represent the model of the current train to be detected. After locating the coordinates of the vehicle number area, it intercepts the corresponding vehicle number area image from the preprocessed image and then recognizes the vehicle number area image, thereby removing the remaining interference information in the image so that only the model of the train to be detected can be accurately and to the greatest extent possible.

[0069] Once the model of the train to be inspected is identified, the system retrieves the corresponding vehicle model characteristic parameters from the database based on the model. These parameters include the vehicle height, high-voltage equipment installation coordinates, and high-voltage equipment height parameters. Based on the vehicle model characteristic parameters, a motion command is generated to control the rotation and adjustment of the servo unit. This command is then sent to the servo unit, causing it to rotate the acoustic sensor to the position corresponding to the high-voltage equipment to be inspected.

[0070] According to an embodiment of the present disclosure, after receiving an appearance image of a train to be inspected, the processing module first performs basic image processing on the appearance image to restore image properties such as resolution to obtain a preprocessed appearance image. An image recognition model is then used to locate and identify the vehicle number region in the preprocessed appearance image, and a portion of the image containing only the vehicle number data is extracted, thereby shielding and removing interference from the remaining image data and retaining only the image data region containing the information to be identified and extracted. The vehicle number region image is then used to identify and extract the model of the train to be inspected, obtaining the specific model of the train to be inspected. Based on the model, vehicle characteristic parameters that match the model and contain characteristic information such as the location of high-voltage equipment and vehicle height are retrieved from a database. Motion instructions for controlling the adjustment of a servo unit are generated based on the vehicle characteristic parameters. This allows a preliminary visual inspection of the train to be inspected before acoustic inspection of the high-voltage equipment. Feedback from the visual inspection is fed back to the servo unit in the acoustic inspection to adjust the relative positions of the servo unit and the acoustic sensor relative to the high-voltage equipment. Without manual adjustment, the detection and acquisition device is adaptively adjusted for different high-voltage equipment to be inspected, optimizing the layout of each sensor to facilitate the acquisition of accurate sound signals.

[0071] According to an embodiment of the present disclosure, the processing module can also be used to perform image recognition on the vehicle number area image, obtain the train number of the train to be detected, and write the fault detection result into the digital maintenance file corresponding to the train number.

[0072] In addition to the model of the train to be detected, the vehicle number area image may also include image data of the train number of the train to be detected. The processing module can determine the train number of the train to be detected during the process of recognizing the vehicle number area image.

[0073] Each train has a corresponding train number. Based on the train number, a digital maintenance file corresponding to each train can be generated. In the digital maintenance file, the historical maintenance data and current maintenance data of the two trains can be retrieved at will, and relevant subsequent processing of the train can be carried out based on this data.

[0074] According to an embodiment of the present disclosure, a remote monitoring platform may also be provided in the multimodal diagnosis host.

[0075] According to an embodiment of the present disclosure, the remote monitoring platform is communicatively connected to the processing module, and the remote monitoring platform is used to receive a query operation, determine the target train number indicated by the query operation, and send a query request for the target train number to the processing module.

[0076] The remote monitoring platform can be located at the top level of the system architecture and connected to the multimodal diagnostic host in the middle level via Ethernet communication links, allowing operators to complete system-wide control without having to touch the high-voltage equipment. The remote monitoring platform can monitor and display the status of various system components, fence intrusion alarms, manage digital maintenance files, and display diagnostic results for target equipment.

[0077] According to an embodiment of the present disclosure, the processing module can also be used to respond to a query request, determine the target digital maintenance file corresponding to the target train number, and send the target digital maintenance file to the remote monitoring platform so that the remote monitoring platform can display the target digital maintenance file.

[0078] According to the embodiments of the present disclosure, a remote monitoring platform can also be provided in the entire system architecture of fault detection. After each fault detection is performed on the high-voltage equipment of the train, the fault detection result is written into the digital maintenance file corresponding to the train to be detected. When the operator needs to obtain the historical fault detection results of a certain train, the target train number to be queried can be determined through the upstream remote monitoring platform, and a query request can be sent to the downstream processing module. The processing module retrieves the target digital maintenance file and sends it to the remote monitoring platform, so that the remote monitoring platform can present the target digital maintenance file to the operator, so that the operator can intuitively review the detection results and ensure the stability of the link transmission between each sensor and module and platform.

[0079] Figure 2 The following schematically shows a remote monitoring platform to which the embodiments of the present disclosure can be applied.

[0080] like Figure 2 As shown, the remote monitoring platform can display the communication status 201 of each module, the overall control status 202, the servo unit status 203, the acquisition progress 204, the analysis progress 205, the detection information 206, the first visual acquisition module 207, the first second visual acquisition module 208 and the second second visual acquisition module 209. The overall control status 202 can include start and stop, the acquisition progress 204 can include image acquisition progress and sound acquisition progress, the analysis progress 205 can include image analysis progress and sound analysis progress, the detection information 206 can include the current train information to be detected, the fault detection results and the digital maintenance file, the first visual acquisition module 207 is used to collect the appearance image of the train to be detected, the first second visual acquisition module 208 is used to collect the first environmental image of the train to be detected and the second second visual acquisition module 209 is used to collect the second environmental image of the train to be detected.

[0081] According to an embodiment of the present disclosure, the second visual acquisition module is configured to be set relative to the second target carriage of the train to be inspected, and the second target carriage includes a carriage equipped with high-voltage equipment; the second visual acquisition module is used to acquire an environmental image containing the high-voltage equipment and send the environmental image to the processing module.

[0082] When performing fault detection on the high-voltage equipment of the train, multiple second visual acquisition modules can be set up. Multiple second visual acquisition modules can be set up at the guardrails of the top platform of the maintenance depot at the stop of the middle train and the tail train, for example, at the guardrails of the top platform of the maintenance depot at the stop of the 03 and 06 carriages of the train.

[0083] The train with the high-voltage equipment to be inspected is photographed by multiple second vision acquisition modules, so that the processor can detect the current fault detection environment of the high-voltage equipment based on the multiple environmental images photographed by the multiple second vision acquisition modules.

[0084] According to an embodiment of the present disclosure, the processing module can also be used to perform environmental safety detection based on environmental images to obtain detection results, and when the detection results indicate that there are no abnormal objects around the high-voltage equipment, execute motion instructions to drive the servo unit to move and change the spatial distribution of multiple acoustic sensors.

[0085] After the second visual acquisition module captures the environmental image, the processing module performs environmental safety detection on the environmental image related to fence intrusion detection. The input environmental image is analyzed through the environmental assessment model to determine whether there are foreign objects blocking or human activities around the train and high-voltage equipment to be inspected, so as to avoid detection accidents caused by external interference during the detection process.

[0086] According to an embodiment of the present disclosure, the processing module may also be configured to feed back a warning signal to the remote monitoring platform and stop the execution of the motion instruction when the detection result indicates that there are abnormal objects around the high-voltage equipment.

[0087] In the case of abnormal objects, timely feedback is required to the operator through the remote monitoring platform so that the operator can control and eliminate the relevant abnormal objects. When the detection results indicate that there are no abnormal objects around the high-voltage equipment, the appearance image collected by the first visual acquisition module is processed normally to generate and send motion instructions to the servo unit.

[0088] According to the embodiments of the present disclosure, while the first visual acquisition module is used to acquire the appearance image of the train to be detected, the second visual acquisition module is used to acquire the environmental image around the train to be detected, and then the processing module is used to perform real-time fence intrusion detection on the environmental image. When an abnormal object appears around the train to be detected, the generation and issuance of motion instructions are suspended in time, and the relevant personnel are notified of the appearance of the abnormal object. Only when no abnormal object appears around the train to be detected will the motion instruction be normally issued to the servo unit, thereby avoiding abnormal events in the environment to be detected, improving the safety assessment of the detection environment, and thus ensuring the safety of operators and detection equipment during the diagnosis process.

[0089] Figure 3 A schematic diagram of a fault detection device for high-voltage equipment of a rail transit train to which another embodiment of the present disclosure can be applied is shown schematically.

[0090] like Figure 3 As shown, when it is necessary to perform fault detection on the high-voltage equipment on the train to be inspected, the train to be inspected needs to be stopped at a predetermined train stop position first, and then the first visual acquisition module located at the railing position of the 01 carriage is used to collect the appearance image of the train to be inspected, and the second visual acquisition module located at the railing position of the 03 and 06 carriages is used to collect the environmental image of the train to be inspected, and then both the appearance image and the environmental image are transmitted to the processing module for generating motion instructions and judging fence intrusion detection. When it is judged that there are no abnormal things, motion instructions are issued to servo unit 1 and servo unit 2. Servo unit 1 and servo unit 2 respond to the motion instructions with acoustic sensors for adaptive adjustment. The acoustic sensor collects the sound signal of the high-voltage equipment under normal operation and transmits the sound signal to the processing module. The processing module performs fault detection processing on the sound signal to obtain the fault detection result, and writes the fault detection result into the digital maintenance file corresponding to the train number, so that the operator can obtain the current fault detection result and the historical fault detection result from the upper remote monitoring platform.

[0091] Figure 4 A flowchart of a method for detecting faults in high-voltage equipment of a rail transit train to which an embodiment of the present disclosure can be applied is schematically shown.

[0092] like Figure 4 As shown, the fault detection method for high-voltage equipment of a rail transit train includes operations S410 to S420.

[0093] In operation S410 , sound signals collected by a plurality of acoustic sensors are acquired, where the plurality of acoustic sensors are distributed around a high-voltage device of a train to be inspected.

[0094] In operation S420 , fault detection is performed based on the plurality of sound signals to obtain a fault detection result.

[0095] According to the embodiments of the present disclosure, when fault detection is performed on high-voltage equipment, a processing module is used to analyze and process the sound signals collected by multiple acoustic sensors uniformly surrounding the high-voltage equipment for fault detection, so as to obtain fault detection results. This enables accurate detection results of the high-voltage equipment of the train to be obtained under non-contact detection, thereby improving the detection efficiency and accuracy of the high-voltage equipment and the safety of the detection equipment and personnel. At the same time, it reduces dependence on the internal hardware performance of the train, reduces the memory resources consumed by fault detection, and improves the adaptability range and application environment of fault detection.

[0096] Figure 5 A flowchart of a method for detecting faults in high-voltage equipment of a rail transit train to which another embodiment of the present disclosure can be applied is schematically shown.

[0097] like Figure 5 As shown, start S501, start the remote monitoring platform and processing module S502, perform link self-check S503 between the remote monitoring platform, processing module and sensor, the first visual acquisition module acquires the appearance image of the train to be detected S504, matches or establishes a digital maintenance file S505, the second visual acquisition module acquires the environmental image of the train to be detected S506, generates motion instructions and controls the servo unit to adjust the position S507, the acoustic sensor acquires sound signals S508, the processing module performs fault detection on the sound signals S509, the servo unit returns to the standby position S510, the remote monitoring platform displays the fault detection results S511, and ends S512.

[0098] Figure 6 A block diagram of an electronic device for a method for detecting a fault of high-voltage equipment of a rail transit train according to an embodiment of the present disclosure is schematically shown.

[0099] like Figure 6 As shown, the electronic device according to an embodiment of the present disclosure includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage portion 608 into a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0100] Various programs and data required for the operation of the electronic device are stored in RAM 603. The processor 601, ROM 602, and RAM 603 are connected to each other via a bus 604. The processor 601 executes the programs in ROM 602 and / or RAM 603 to perform various operations of the method flow according to the embodiment of the present disclosure. It should be noted that the programs may also be stored in one or more memories other than ROM 602 and RAM 603. The processor 601 may also execute the programs stored in the one or more memories to perform various operations of the method flow according to the embodiment of the present disclosure.

[0101] According to an embodiment of the present disclosure, the electronic device may further include an input / output (I / O) interface 605, which is also connected to the bus 604. The electronic device may further include one or more of the following components connected to the I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including devices such as a cathode ray tube (CRT), liquid crystal display (LCD), and speakers; a storage section 608 including a hard disk; and a communication section 609 including a network interface card such as a LAN card or modem. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. Removable media 611, such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed in the drive 610 as needed, so that computer programs read from the removable media can be installed in the storage section 608 as needed.

[0102] According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the processor 601, the above-mentioned functions defined in the system of the embodiment of the present disclosure are executed. According to an embodiment of the present disclosure, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0103] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not be incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, and when executed, implements the method according to the embodiments of the present disclosure.

[0104] According to embodiments of the present disclosure, a computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples include, but are not limited to, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may 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.

[0105] For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the ROM 602 and / or the RAM 603 described above and / or one or more memories other than the ROM 602 and the RAM 603 .

[0106] An embodiment of the present disclosure also includes a computer program product, which includes a computer program, which contains program code for executing the method provided by the embodiment of the present disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the fault detection method for high-voltage equipment of a rail transit train provided by the embodiment of the present disclosure.

[0107] When the computer program is executed by the processor 601, the above functions defined in the system / device of the embodiment of the present disclosure are performed. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0108] In one embodiment, the computer program may be stored on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal on a network medium, downloaded and installed via the communication portion 609, and / or installed from a removable medium 611. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0109] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computer 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 computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect via the Internet).

[0110] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, as well as the combination of boxes in the block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified functions or operations, or can be implemented using a combination of dedicated hardware and computer instructions. It will be understood by those skilled in the art that the features described in the various embodiments of the present disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features described in the various embodiments of the present disclosure may be combined and / or coupled in various ways, and all of these combinations and / or couplings fall within the scope of the present disclosure.

[0111] The above describes the embodiments of the present disclosure. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although each embodiment has been described separately above, this does not mean that the measures in each embodiment cannot be advantageously used in combination. Without departing from the scope of the present disclosure, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A fault detection device for high-voltage equipment of a rail transit train, comprising: An audio acquisition module includes a servo unit and a plurality of acoustic sensors, wherein the plurality of acoustic sensors are provided on the servo unit, and the servo unit is configured to move under the control of a motion instruction so that the plurality of acoustic sensors surround the high-voltage equipment distributed on the train to be inspected; as well as The processing module is used to obtain the sound signals collected by each of the multiple acoustic sensors when the high-voltage equipment of the train to be detected is in an operating state, and perform fault detection based on the multiple sound signals to obtain a fault detection result.

2. The device according to claim 1, wherein The processing module is used to extract audio features from the multiple sound signals respectively to obtain multiple Mel-spectrograms, and use the mean aggregation mechanism and the maximum probability mechanism to make decisions based on the multiple Mel-spectrograms to obtain fault detection results.

3. The device according to claim 2, wherein The processing module is used to downsample the sound signal to obtain a target sound signal, perform time domain segmentation on the target sound signal to obtain multiple time domain frame signals, perform Fourier transform on the multiple time domain frame signals to obtain multiple frequency domain frame signals, and use a Mel filter group to map the multiple frequency domain frame signals to obtain the logarithmic Mel spectrum diagram.

4. The device according to claim 2, wherein The processing module is used to use the mean aggregation mechanism to perform global mean calculation on the fault probability distribution output from each time window of the multiple logarithmic Mel-spectrograms, determine the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment, and use the maximum probability mechanism to accumulate the fault categories with the highest fault probability in each time window of the multiple logarithmic Mel-spectrograms to determine the number of occurrences of each of the multiple types of faults, and obtain the fault detection result based on the probability values ​​of each of the multiple types of faults occurring in the high-voltage equipment and the number of occurrences of each of the multiple types of faults.

5. The apparatus according to claim 1, further comprising: The first visual acquisition module is configured to be set relative to the first target carriage of the train to be detected. The first visual acquisition module is used to collect the appearance image of the train to be detected and send the appearance image to the processing module.

6. The device according to claim 5, wherein The processing module is further configured to determine the model of the train to be detected based on the appearance image, obtain vehicle model characteristic parameters corresponding to the model of the train to be detected from a database, and generate motion instructions for the servo unit based on the vehicle model characteristic parameters.

7. The device according to claim 6, wherein The processing module is used to perform dynamic range compression and distortion correction on the appearance image to obtain a preprocessed appearance image, use an image recognition model to identify the vehicle number area of ​​the preprocessed appearance image to obtain the vehicle number area coordinates, based on the vehicle number area coordinates, intercept the vehicle number area image from the preprocessed appearance image, and perform image recognition on the vehicle number area image to determine the model of the train to be detected.

8. The device according to claim 7, wherein The processing module is further configured to perform image recognition on the vehicle number area image to obtain the train number of the train to be detected, and write the fault detection result into a digital maintenance file corresponding to the train number.

9. The apparatus according to claim 8, further comprising: a remote monitoring platform, communicatively connected to the processing module, the remote monitoring platform being configured to receive a query operation, determine a target train number indicated by the query operation, and send a query request for the target train number to the processing module; The processing module is also used to respond to the query request, determine the target digital maintenance file corresponding to the target train number, and send the target digital maintenance file to the remote monitoring platform so that the remote monitoring platform can display the target digital maintenance file.

10. The apparatus according to claim 1, further comprising: a second visual acquisition module, configured to be disposed relative to a second target carriage of the train to be inspected, wherein the second target carriage includes a carriage in which the high-voltage equipment is disposed; The second visual acquisition module is used to acquire an environmental image containing the high-voltage equipment and send the environmental image to the processing module.

11. The device according to claim 10, wherein The processing module is also used to perform environmental safety detection based on the environmental image to obtain a detection result, and when the detection result indicates that there are no abnormal objects around the high-voltage equipment, execute the motion instruction to drive the servo unit to move and change the spatial distribution of the multiple acoustic sensors.

12. The device according to claim 11, wherein The processing module is further configured to feed back a warning signal to a remote monitoring platform and stop the execution of the motion instruction when the detection result indicates that an abnormal object exists around the high-voltage equipment.

13. A method for detecting a fault in a high-voltage device of a rail transit train, applied to the apparatus for detecting a fault in a high-voltage device of a rail transit train according to any one of claims 1 to 12, the method comprising: Acquiring sound signals collected by a plurality of acoustic sensors, wherein the plurality of acoustic sensors are distributed around a high-voltage device of a train to be inspected; as well as Fault detection is performed based on the plurality of sound signals to obtain a fault detection result.

14. An electronic device comprising: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method of claim 13.

15. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, causes the processor to implement the method of claim 13.

16. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to claim 13.

Images