Motor fault detection method and device and computer readable storage medium

Abstract

The invention provides a motor fault detection method and device and a computer readable storage medium. The motor fault detection method comprises the following steps: acquiring a vibration signal, a sound wave signal and thermal image information in a motor operation process; key features are respectively extracted from the vibration signal, the sound wave signal and the thermal image information. Matching and fusing the key features of the vibration signal, the key features of the sound wave signal and the key features of the thermal image information to obtain a comprehensive feature vector; and analyzing the comprehensive feature vector, and outputting a fault position and a fault type of the motor. And the motor fault detection is more accurate.

Description

Technical Field

[0001] This application relates to the field of motor fault detection, and in particular to a method, apparatus and computer-readable storage medium for motor fault detection. Background Technology

[0002] As the core drive component of the integrated test platform, the main pump motor's performance and reliability directly affect the overall performance and service life. During long-term operation, due to factors such as friction, vibration, and thermal stress, the internal mechanical components (such as bearings, gears, and windings) are prone to wear, leading to a decline in motor performance or even failure.

[0003] In related technologies, vibration signals generated during motor operation can be monitored and their spectrum analyzed to identify mechanical wear characteristics. However, vibration signals are easily interfered with by vibrations and noise from other mechanical equipment in the surrounding environment, leading to a decrease in the signal-to-noise ratio and affecting the accuracy and reliability of the detection. Furthermore, the vibration signals generated during motor operation are complex and variable, containing multiple frequency components, making the separation and identification of characteristic signals difficult. Moreover, vibration signals may not cover all critical parts of the motor, resulting in poor detection comprehensiveness and hindering accurate detection of motor faults. Summary of the Invention

[0004] This application provides a method, apparatus, and computer-readable storage medium for motor fault detection, which enables more accurate motor fault detection.

[0005] This application provides a method for detecting motor faults, including: The vibration signal, sound wave signal, and thermal image information of the motor during operation are acquired. Key features are extracted from the vibration signal, the acoustic signal, and the thermal image information, respectively. The key features of the vibration signal, the key features of the acoustic signal, and the key features of the thermal image information are matched and fused to obtain a comprehensive feature vector. The comprehensive feature vector is analyzed, and the fault location and fault type of the motor are output.

[0006] Furthermore, acquiring the vibration signal during the operation of the motor includes: Multiple vibration signals collected by multiple acceleration sensors during the operation of the motor are acquired; wherein the multiple acceleration sensors are at least installed in the bearings, rotor and housing of the motor; Key features are extracted from the vibration signal, including: Key features are extracted from the multiple vibration signals.

[0007] Furthermore, the extraction of key features from the multiple vibration signals includes: Bandpass filtering is performed on multiple vibration signals; Spectral analysis was performed on the filtered vibration signals to obtain multiple vibration spectral signals. Key features are extracted from multiple vibration spectrum signals.

[0008] Further, acquiring the acoustic signal during the operation of the motor includes: The system acquires multiple sound wave signals collected by multiple microphones during the operation of the motor; wherein the multiple microphones are at least located at the front end, rear end, and housing of the motor. Extracting key features from the acoustic signal includes: Key features are extracted from the multiple acoustic signals.

[0009] Furthermore, the extraction of key features from the plurality of acoustic signals includes: Bandpass filtering is performed on multiple of the aforementioned acoustic signals; Spectral analysis was performed on the filtered acoustic signals to obtain multiple acoustic spectral signals. Key features are extracted from multiple acoustic spectrum signals.

[0010] Furthermore, acquiring the thermal image information during the operation of the motor includes: Multiple thermal image information collected by multiple infrared thermal imagers during the operation of the motor is acquired; wherein the multiple infrared thermal imagers are arranged around the motor and facing the motor; Key features are extracted from the thermal image information, including: Key features are extracted from multiple thermal image data.

[0011] Furthermore, the extraction of key features from the multiple thermal image information includes: Preprocessing is performed on multiple thermal image information; Key features are extracted from the preprocessed thermal image information.

[0012] Furthermore, the motor fault detection method also includes: The environmental information of the area where the motor is located is collected by the environmental detection module; If the environmental information exceeds a preset range, the environmental control device is controlled to adjust the environmental information of the area where the motor is located; the environmental information includes at least one of temperature, humidity and noise.

[0013] This application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the motor fault detection method as described in any of the above embodiments.

[0014] This application provides a motor fault detection device, including a control system. The control system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the motor fault detection method as described in any of the above embodiments.

[0015] The motor fault detection method provided in this application can acquire vibration signals, acoustic signals, and thermal image information during motor operation. By matching and fusing key features of the vibration signals, acoustic signals, and thermal image information, a comprehensive feature vector is obtained. The comprehensive feature vector is analyzed, and the fault location and fault type of the motor are output. Thus, by matching and fusing key features of the vibration signals, acoustic signals, and thermal image information, the comprehensiveness of motor fault detection is improved, making motor fault detection more accurate.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] Figure 1 The diagram shown is a structural diagram of one embodiment of the motor fault detection device of this application; Figure 2 The diagram shown is a flowchart of one embodiment of the motor fault detection method of this application; Figure 3 The diagram shown is a flowchart of a sub-implementation of the motor fault detection method of this application; Figure 4 The diagram shown is a module schematic of the control system of the motor fault detection device of this application. Detailed Implementation

[0019] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0020] This application provides a motor fault detection method, a computer-readable storage medium, and a motor fault detection device. The motor fault detection method provided by this application includes acquiring vibration signals, acoustic signals, and thermal image information during motor operation. Key features are extracted from the vibration signals, acoustic signals, and thermal image information respectively. The key features of the vibration signals, acoustic signals, and thermal image information are matched and fused to obtain a comprehensive feature vector. The comprehensive feature vector is analyzed, and the fault location and fault type of the motor are output.

[0021] This application provides a motor fault detection method, a computer-readable storage medium, and a motor fault detection device. By matching and fusing key features of vibration signals, key features of acoustic signals, and key features of thermal image information, the comprehensiveness of motor fault detection is improved, making motor fault detection more accurate.

[0022] Figure 1 The diagram shown is a structural diagram of one embodiment of the motor fault detection device 10 of this application. Figure 2 The diagram shown is a flowchart of one embodiment of the motor fault detection method of this application. See also... Figure 1 and Figure 2 As shown, the motor fault detection method includes steps S1 to S4.

[0023] Step S1: Acquire vibration signals, acoustic signals, and thermal image information during the operation of motor 11. This can be achieved by acquiring vibration signals collected by accelerometer 12, acoustic signals collected by microphone 13, and thermal image information collected by infrared thermal imager 14 during the operation of motor 11.

[0024] In one embodiment, after a preset time elapsed since the motor 11 starts, the accelerometer 12, microphone 13, and infrared thermal imager 14 are simultaneously activated, enabling them to acquire data concurrently. This ensures the temporal consistency of data acquired by different detection methods, providing an accurate and synchronized data foundation for subsequent matching and fusion of key features. A unified time reference can be provided by a time synchronization module to calibrate the vibration detection module, acoustic detection module, and thermal detection module. A synchronization trigger signal is generated and transmitted to the vibration detection module, acoustic detection module, and thermal detection module respectively through a trigger signal distributor to ensure simultaneous data acquisition. The acquired data is temporarily stored through a data buffer to coordinate the data transmission rate. The vibration detection module includes an accelerometer 12, the acoustic detection module includes a microphone 13, and the thermal detection module includes an infrared thermal imager 14.

[0025] Step S2: Extract key features from vibration signals, sound wave signals, and thermal image information respectively. Key features can be extracted from vibration signals, sound wave signals, and thermal image information.

[0026] Step S3: Match and fuse the key features of the vibration signal, the key features of the acoustic signal, and the key features of the thermal image information to obtain a comprehensive feature vector.

[0027] Step S4: Analyze the comprehensive feature vector and output the fault location and fault type of motor 11. The comprehensive feature vector can be analyzed using a Transformer-based motor fault diagnosis model to determine the fault location and fault type of motor 11. Specifically, the comprehensive feature vector can be input into the trained motor fault diagnosis model to perform fault diagnosis of motor 11, identifying the fault location and fault type end-to-end, and assessing the motor's health status.

[0028] In one embodiment, a motor fault diagnosis model can be trained by inputting training samples into a convolutional neural network. The training samples are labeled multimodal information fusion maps. The raw materials can be processed to obtain the training samples. A supervised learning dataset can be manually constructed by collecting vibration signals, acoustic signals, and thermal image information from faulty and normal motors, labeling the fault location and fault type of the faulty motor as required indicators, and using this as the raw material. The dataset size needs to be greater than 10k. The connection weights and biases of the neural network can be continuously updated through training until the neural network meets the requirements, thus obtaining the motor fault diagnosis model.

[0029] The motor fault detection method provided in this application can acquire vibration signals, acoustic signals, and thermal image information during the operation of motor 11. By matching and fusing key features of the vibration signals, acoustic signals, and thermal image information, a comprehensive feature vector is obtained. The comprehensive feature vector is analyzed, and the fault location and fault type of motor 11 are output. Thus, by matching and fusing key features of vibration signals, acoustic signals, and thermal image information, the comprehensiveness of motor fault detection is improved, making motor fault detection more accurate.

[0030] The motor fault detection method provided in this application can integrate vibration signals, sound wave signals and thermal image information to achieve a comprehensive and accurate assessment of motor mechanical wear.

[0031] In one embodiment, the waveforms and spectrum diagrams of the vibration and acoustic signals, thermal imaging images, and comprehensive analysis results can be displayed on the screen. This makes the display more intuitive, user-friendly, and easy to operate.

[0032] In one embodiment, a detection report can be generated and stored in a database through a report generation module. The motor fault detection device 10 can realize functions such as data visualization, alarm prompts, historical data playback, and report generation to provide decision support for users.

[0033] In one embodiment, acquiring vibration signals during the operation of the motor 11 includes acquiring multiple vibration signals collected by multiple acceleration sensors during the operation of the motor 11. The multiple acceleration sensors 12 are at least disposed on the bearings, rotor, and housing of the motor 11. The acceleration sensors 12 can monitor the vibration signals generated by the motor 11 during operation in real time. The multiple acceleration sensors 12 can be at least disposed on the bearings, rotor ends, and housing of the motor 11. At least two acceleration sensors 12 can be disposed on the bearings of the motor 11, at least two acceleration sensors 12 can be disposed on the rotor ends of the motor 11, and at least two acceleration sensors 12 can be disposed on the housing of the motor 11. The placement of the acceleration sensors 12 needs to avoid core components to prevent interference with the operation of the motor 11. This ensures that the acquired vibration signals are more comprehensive.

[0034] Extracting key features from vibration signals includes extracting key features from multiple vibration signals. Extracting key features from multiple vibration signals results in more accurate extraction.

[0035] In one embodiment, key features are extracted from multiple vibration signals, including: Bandpass filtering can be applied to multiple vibration signals. By using a filter to apply bandpass filtering to multiple vibration signals, noise in the vibration signals can be removed, and the vibration signals can be effectively distinguished from external interference, thereby ensuring the accuracy of the vibration signals.

[0036] Multiple vibration signals after filtering are subjected to spectral analysis to obtain multiple vibration spectrum signals. Multiple vibration spectrum signals can be obtained through signal analysis techniques such as Fast Fourier Transform (FFT). Complex vibration signals can be decomposed into different frequency components, thereby identifying abnormal conditions during motor operation and determining the specific type of fault, such as imbalance, looseness, or bearing failure.

[0037] Key features are extracted from multiple vibration spectrum signals. In one embodiment, the current multiple vibration spectrum signals can be compared with the vibration spectrum signals during normal operation of the motor 11 to identify characteristic frequencies and amplitude changes in the current multiple vibration spectrum signals. Characteristic frequencies include inner race frequency, outer race frequency, rolling element frequency, gear meshing frequency, imbalance frequency, loosening frequency, etc., which correspond to different fault characteristics of various components of the motor 11. In this way, by monitoring the vibration signals generated by the motor 11 during operation in real time, the fault location and fault type of the motor 11 can be determined.

[0038] In one embodiment, acquiring acoustic signals during the operation of the motor 11 includes acquiring multiple acoustic signals collected by multiple microphones 13 during the operation of the motor 11. The multiple microphones 13 are at least located at the front end, rear end, and housing of the motor 11. The multiple microphones 13 may be located at least around the front end, rear end, and housing of the motor 11. The multiple microphones 13 may be located at least in the areas near bearings, gears, and rotors within the front end, rear end, and housing of the motor 11, thus providing complete acoustic information about the operating status of various components of the motor 11 while avoiding distortion or omission of acoustic signals that may occur due to a single location. The microphones 13 can capture acoustic signals generated by the motor 11 in real time during operation.

[0039] Extracting key features from acoustic signals includes extracting key features from multiple acoustic signals. Extracting key features from multiple acoustic signals results in more accurate extraction.

[0040] In one embodiment, extracting key features from multiple acoustic signals includes: Bandpass filtering is applied to multiple acoustic signals. By using a filter to bandpass filter multiple acoustic signals, noise in the acoustic signals can be removed, improving the quality of the acoustic signals. A specific frequency range can be set to retain only the relevant frequency signals generated during the operation of motor 11, while removing high-frequency noise and low-frequency interference signals.

[0041] Multiple filtered acoustic signals are subjected to spectral analysis to obtain multiple acoustic spectrum signals. These signals can be obtained using signal analysis techniques such as the Fast Fourier Transform (FFT). The Fast Fourier Transform converts the acoustic signal into a spectrum.

[0042] Key features are extracted from multiple acoustic spectrum signals. In one embodiment, the current multiple acoustic spectrum signals can be compared with multiple acoustic spectrum signals generated during the normal operation of the motor 11 to analyze abnormal acoustic spectra in the current multiple acoustic spectrum signals. Acoustic detection has the advantage of being non-contact and does not affect the normal operation of the motor 11. Compared with vibration signals, acoustic signals have a wider propagation range, making it easier to capture the sound of internal mechanical wear in the motor 11. In addition, acoustic detection exhibits high sensitivity to certain faults, such as poor gear meshing and bearing defects. Thus, by monitoring the acoustic signals generated by the motor 11 in real time during operation, the location and type of fault in the motor 11 can be determined.

[0043] In one embodiment, acquiring thermal image information during the operation of the motor 11 includes acquiring multiple thermal image information collected by multiple infrared thermal imagers 14 during the operation of the motor 11. The multiple infrared thermal imagers 14 are positioned around the motor 11 and facing it. The infrared thermal imagers 14 can be positioned at least above the motor 11 and can be arranged in four directions around the motor 11. This ensures coverage of all surfaces of the motor 11, capturing comprehensive temperature data. The infrared thermal imagers 14 are used to detect the outer surface of the motor 11, bearings, gears, and other high-temperature areas. The infrared thermal imagers 14 focus the infrared radiation emitted from the surface of the motor 11 onto the surface of an infrared detector through an infrared lens. The infrared detector converts the infrared radiation into a thermal signal and further into an electrical signal.

[0044] Extracting key features from thermal image information includes extracting key features from multiple thermal image sources. Extracting key features from multiple thermal image sources results in more accurate extraction.

[0045] In one embodiment, extracting key features from multiple thermal image data includes preprocessing the multiple thermal image data. This preprocessing may include denoising, image enhancement, and image cropping of the thermal image data.

[0046] Key features are extracted from multiple preprocessed thermal images. In one embodiment, the temperature distribution in the thermal images can be analyzed, and temperature anomalies and hot spots can be identified based on preset temperature thresholds. These anomalies and hot spots are then labeled to provide a basis for fault diagnosis. Thus, by analyzing the temperature distribution in multiple thermal images of the motor 11, localized abnormal temperature increases caused by friction, wear, etc., can be identified. Localized abnormal temperature increases are usually signals of friction or wear of mechanical parts, and may indicate an impending fault. Therefore, the location and type of fault in the motor 11 can be determined using multiple thermal images.

[0047] The infrared thermal imager 14 can be used to detect the temperature distribution of the motor 11 and identify localized abnormal temperature increases caused by friction and wear. The infrared thermal imager 14 has the advantages of being intuitive and fast, and can be used in high-temperature or hazardous environments. Compared to vibration signal analysis, the infrared thermal imager 14 can provide comprehensive temperature information of the motor 11, visually displaying characteristic locations and areas of abnormal temperature through thermal images. Furthermore, even if the infrared thermal imager 14 is placed outside the motor, it can detect heat conduction caused by internal wear, thereby indirectly reflecting the wear condition of internal mechanical components.

[0048] In this embodiment, multiple detection methods, including vibration analysis, acoustic detection, and infrared thermal imaging, are combined to overcome the limitations of single-mode detection. By matching and fusing key features of vibration signals, key features of acoustic signals, and key features of thermal image information, the fault location and fault type of motor 11 are obtained, significantly improving the detection sensitivity and anti-interference capability, thereby achieving comprehensive and accurate detection of the mechanical wear of motor 11.

[0049] Figure 3 The diagram shown is a flowchart of a sub-implementation of the motor fault detection method of this application. See also... Figure 3 As shown, in one embodiment, the motor fault detection method further includes steps S5 to S6.

[0050] Step S5: Obtain environmental information of the area where motor 11 is located, collected by the environmental detection module. Environmental information includes temperature, humidity, and noise. The temperature of the area where motor 11 is located can be collected in real time using a temperature sensor, the humidity of the area where motor 11 is located can be collected in real time using a humidity sensor, and the noise of the area where motor 11 is located can be collected in real time using a noise microphone.

[0051] Step S6: If the environmental information exceeds the preset range, control the environmental control device to adjust the environmental information of the area where the motor is located. The environmental information includes at least one of temperature, humidity, and noise. This ensures the stability and consistency of the detection environment. Specifically, if the environmental information exceeds the preset range, controlling the environmental control device to adjust the environmental information of the area where the motor 11 is located can, for example, activate the temperature control module, humidity control module, and noise shielding module. In the detection of mechanical wear of the motor 11, data such as vibration signals, sound wave signals, and thermal image information may be affected by environmental factors. For example, if the temperature of the area where the motor 11 is located is too high, it may cause the accelerometer 12 and other sensors to overheat, leading to signal distortion. If the temperature of the area where the motor 11 is located is too low, the accelerometer 12 and other sensors may exhibit a slow response. Adjusting the environmental information of the area where the motor 11 is located, if the environmental information exceeds the preset range, ensures that the sensors are always in a good working state, avoiding measurement errors caused by environmental changes.

[0052] In this embodiment, the specific workflow is as follows: It can activate environmental control and equipment, and control the environmental control equipment to adjust the environmental information of the area where the motor is located, so as to ensure the stability of the detection environment.

[0053] After a preset time elapses after the motor 11 starts, the accelerometer 12, microphone 13, and infrared thermal imager 14 are simultaneously activated, enabling the accelerometer 12, microphone 13, and infrared thermal imager 14 to collect vibration signals, sound wave signals, and thermal image information in real time, respectively.

[0054] Noise is removed from vibration and acoustic signals using filters, and the signals are normalized. Key features are then extracted from the vibration, acoustic, and thermal image information using a feature extraction neural network. Specifically, the feature extraction network of the vibration detection module processes and extracts key features from the vibration signal, the acoustic signal, and the thermal image information.

[0055] The key features of vibration signals, sound signals, and thermal image information are matched and fused to obtain a comprehensive feature vector.

[0056] The comprehensive feature vector is analyzed, and the fault location and fault type of motor 11 are output. The comprehensive feature vector can be analyzed using a Transformer-based motor fault diagnosis model.

[0057] The display screen shows waveforms and spectrum diagrams of vibration and sound signals, thermal imaging images, and comprehensive analysis results. When an abnormality or fault is detected, the alarm module is triggered to prompt the user to take action.

[0058] The report generation module can generate test reports and store them in the database. The test report includes test data, analysis results, and maintenance recommendations. The database stores the test data, analysis results, and generated report, and supports historical data retrieval.

[0059] Figure 4 The diagram shown is a module schematic of the control system 16 of the motor fault detection device 10 of this application. See also... Figure 4As shown in the illustration, this application embodiment also provides a motor fault detection device 10, including a control system 16. The control system 16 includes a memory, a processor 18, and a computer program stored in the memory and executable on the processor 18. When the processor 18 executes the program, it implements the steps of the motor fault detection method described above. The control system 16 may include one or more processors 18 for implementing the motor fault detection method described above.

[0060] In one embodiment, the control system 16 may include a storage medium 19, such as a computer-readable storage medium that stores a program that can be called by the processor 18, and may include a non-volatile storage medium.

[0061] In one embodiment, the control system 16 may include memory 20 and interface 21. In another embodiment, the control system 16 may also include other hardware depending on the specific application.

[0062] The computer-readable storage medium of this application embodiment stores a program thereon, which, when executed by the processor 18, is used to implement the motor fault detection method as described above.

[0063] This application also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the motor fault detection method as described in any of the preceding claims.

[0064] This application also provides a computer program stored in a computer-readable storage medium, for example... Figure 4 The storage medium 19, and when the processor 18 executes the computer program, causes the processor 18 to execute the motor fault detection method described above.

[0065] This application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented using any method or technology. Information may be computer-readable instructions, data structures, program modules, or other data. Examples of computer-readable storage media include, but are not limited to: phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0066] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method of detecting a fault in an electric machine, characterized by, The method comprises the following steps: acquiring vibration signals, sound wave signals and thermal image information during the operation of the motor; extracting key features from the vibration signals, the sound wave signals and the thermal image information respectively; matching and fusing the key features of the vibration signals, the key features of the sound wave signals and the key features of the thermal image information to obtain a comprehensive feature vector; analyzing the comprehensive feature vector and outputting the fault position and the fault type of the motor.

2. The motor fault detection method of claim 1, wherein, The step of acquiring the vibration signals during the operation of the motor comprises the following steps: acquiring a plurality of vibration signals collected by a plurality of acceleration sensors during the operation of the motor; wherein the plurality of acceleration sensors are arranged at least on the bearings, the rotor and the shell of the motor; the step of extracting key features from the vibration signals comprises the following step: extracting key features from the plurality of vibration signals.

3. The motor fault detection method of claim 2, wherein, The step of extracting key features from the plurality of vibration signals comprises the following steps: band-pass filtering the plurality of vibration signals; performing frequency spectrum analysis on the filtered plurality of vibration signals respectively to obtain a plurality of vibration frequency spectrum signals; extracting key features from the plurality of vibration frequency spectrum signals.

4. The motor fault detection method of claim 1, wherein The step of acquiring the sound wave signals during the operation of the motor comprises the following steps: acquiring a plurality of sound wave signals collected by a plurality of microphones during the operation of the motor; wherein the plurality of microphones are arranged at least at the front end, the rear end and the shell of the motor; the step of extracting key features from the sound wave signals comprises the following step: extracting key features from the plurality of sound wave signals.

5. The method of claim 4, wherein, The step of extracting key features from the plurality of sound wave signals comprises the following steps: band-pass filtering the plurality of sound wave signals; performing frequency spectrum analysis on the filtered plurality of sound wave signals respectively to obtain a plurality of sound wave frequency spectrum signals; extracting key features from the plurality of sound wave frequency spectrum signals.

6. The method of claim 1, wherein, The step of acquiring the thermal image information during the operation of the motor comprises the following steps: acquiring a plurality of thermal image information collected by a plurality of infrared thermal imagers during the operation of the motor; wherein the plurality of infrared thermal imagers are arranged around the motor and face the motor; the step of extracting key features from the thermal image information comprises the following step: extracting key features from the plurality of thermal image information.

7. The method of claim 6, wherein, The step of extracting key features from the plurality of thermal image information comprises the following steps: preprocessing the plurality of thermal image information; extracting key features from the preprocessed plurality of thermal image information.

8. The method of claim 1, wherein, The motor fault detection method further comprises the following steps: acquiring environmental information of the area where the motor is located collected by an environmental detection module; if the environmental information exceeds a preset range, controlling an environmental control device to adjust the environmental information of the area where the motor is located; the environmental information comprises at least one of temperature, humidity and noise.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The program is executed by a processor to implement the steps of the motor fault detection method according to any one of claims 1-8.

10. An electric machine fault detection apparatus, characterized by, The control system comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the program to implement the steps of the motor fault detection method according to any one of claims 1-8.

Images