Fault diagnosis method and device for bearing of motor shaft, equipment and medium

Abstract

The invention discloses a fault diagnosis method and device for a bearing of a motor shaft, equipment and a medium, and relates to the technical field of computers, and the method comprises the steps: obtaining a radio frequency signal generated by electric spark discharge of the bearing of the motor shaft, the pulse density and peak factor of the radio frequency signal, and the energy ratio of the radio frequency signal in a target frequency band; generating a baseband signal of a target frequency band according to the radio frequency signal, and determining the duration of a pulse in a preset proportion according to the radio frequency signal; and diagnosing the fault of the bearing of the motor shaft according to the pulse density, the peak factor, the energy ratio, the baseband signal and the duration. According to the method, the bearing fault of the motor shaft can be diagnosed under the preconditions of low cost and high reliability.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, equipment, and medium for diagnosing a fault of a motor shaft bearing. Background Art

[0002] As a key transmission component, the bearings of the motor shafts in new energy vehicle transmissions are subject to high safety and reliability requirements. With the advancement of industry technology, new energy vehicle powertrains are increasingly moving towards high-voltage platforms, high speeds, lightweight construction, and high loads. These technological upgrades have exacerbated the harsh operating environment of motor shaft bearings, dramatically increasing the risk of failures such as electrical corrosion, poor lubrication, and excessive fatigue damage. Currently, bearing monitoring often utilizes a combination of vibration monitoring and analysis, acoustic monitoring, and temperature detection.

[0003] Among them, the vibration analysis method: requires the installation of an acceleration sensor, the signal-to-noise ratio is low at high speeds or complex working conditions, the vibration signal is easily interfered with by external excitation sources, and the cost of sensors and data acquisition and analysis instruments is extremely high; acoustic monitoring: is easily interfered with by environmental noise, and is not suitable for the noisy environment of new energy power cabins, and acoustic monitoring requires the use of acoustic sensors and dedicated data acquisition and analysis instruments, which are extremely expensive; temperature detection: has obvious hysteresis, regardless of whether it is invasive temperature detection, the hysteresis cannot be effectively eliminated, and temperature detection cannot be used for early warning.

[0004] Therefore, it is necessary to develop a low-cost and high-reliability bearing fault diagnosis method for motor shafts. Summary of the Invention

[0005] The present application provides a method, device, equipment and medium for diagnosing a fault of a motor shaft bearing, which can diagnose a motor shaft bearing fault under the premise of low cost and high reliability.

[0006] To achieve the above objectives, this application adopts the following technical solutions: In a first aspect, the present application provides a method for diagnosing a fault of a bearing of a motor shaft, the method comprising: Acquire a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft, as well as a pulse density, a crest factor, and an energy ratio of the radio frequency signal in a target frequency band; generating a baseband signal of a target frequency band according to the radio frequency signal, and determining a duration of a pulse of a preset proportion according to the radio frequency signal; A fault of a bearing of the motor shaft is diagnosed according to the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration.

[0007] Optionally, diagnosing a fault of a bearing of the motor shaft according to the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration includes: Vectorizing the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration to obtain a multi-dimensional radio frequency vector; Inputting the multi-dimensional radio frequency vector into a fault diagnosis model of a motor bearing to obtain a fault diagnosis result; The motor bearing fault diagnosis model is trained based on sample data, wherein the sample data includes sample features and sample labels, and the sample features include sample pulse density, sample peak factor, sample energy ratio, sample baseband signal and sample duration.

[0008] Optionally, the method further includes: If the fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty, the fault type of the motor shaft bearing is determined.

[0009] Optionally, determining the fault type of the bearing of the motor shaft includes: Acquiring the bearing temperature of the motor shaft and the shaft voltage at both ends of the bearing of the motor shaft; If the bearing temperature of the motor shaft is greater than a preset temperature, the fault type is determined to be lubrication failure; or if the shaft voltage across the bearing of the motor shaft is greater than a preset voltage, the fault type is determined to be insulation layer failure of the motor shaft.

[0010] Optionally, obtaining a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft includes: Acquire a first frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a Rogowski coil array, and acquire a second frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a logarithmic periodic antenna; calculating an average frequency of the first frequency and the second frequency; If the average frequency is lower than a preset frequency, a Rogowski coil array is used to obtain a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft; If the average frequency is greater than the preset frequency, a log-periodic antenna is used to acquire a radio frequency signal generated by the spark discharge of the bearing of the motor shaft.

[0011] Optionally, generating a baseband signal of a target frequency band according to the radio frequency signal includes: Adjust the frequency of the local oscillator to the center frequency of the target frequency band; Determining an in-phase component and a quadrature component of an orthogonal local oscillator signal according to the center frequency and the radio frequency signal; The sum of the squares of the in-phase component and the quadrature component of the quadrature local oscillator signal is determined as the baseband signal of the target frequency band.

[0012] Optionally, the method further includes: If the fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty, a prompt is given through the vehicle computer system.

[0013] In a second aspect, the present application provides a fault diagnosis device for a bearing of a motor shaft, characterized in that the device comprises: an acquisition module, configured to acquire a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft, as well as a pulse density, a crest factor, and an energy ratio of the radio frequency signal in a target frequency band; a determination module, configured to generate a baseband signal of a target frequency band based on the radio frequency signal, and determine a duration of pulses of a preset proportion based on the radio frequency signal; A diagnostic module is used to diagnose a fault of the bearing of the motor shaft according to the pulse density, the peak factor, the energy ratio, the baseband signal and the duration.

[0014] In a third aspect, the present application provides a computing device, including a memory and a processor; One or more computer programs are stored in the memory, and the one or more computer programs include instructions; when the instructions are executed by the processor, the computing device executes the method as described in the first aspect.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program for executing the method described in the first aspect.

[0016] It can be seen from the above technical solution that this application has at least the following beneficial effects: The motor shaft bearing fault diagnosis method proposed in this application effectively solves the lag problem of vibration analysis, temperature detection, and acoustic detection methods. Traditional methods rely on the development of the fault to a certain stage (such as increased vibration, significantly increased temperature, and significant noise) before detection, while this method uses the radio frequency signal generated by the electric spark discharge of the bearing to capture key information at the early stage of the fault. By obtaining the pulse density, peak factor, and target frequency band energy ratio of the radio frequency signal, combined with the generated baseband signal and pulse duration for comprehensive analysis, the bearing status can be reflected in real time and sensitively. This multi-dimensional diagnostic method based on radio frequency signal characteristics breaks through the traditional method's dependence on the degree of fault development, realizes the advanced identification of early faults, greatly improves the timeliness and accuracy of diagnosis, provides more efficient early warning for motor system maintenance, reduces the risk of downtime due to delayed fault discovery, and ensures stable equipment operation. It has significant technical advantages and practical value.

[0017] It should be understood that the description of technical features, technical solutions, beneficial effects or similar language in this application does not imply that all features and advantages can be realized in any single embodiment. On the contrary, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution or beneficial effect is included in at least one embodiment. Therefore, the description of a technical feature, technical solution or beneficial effect in this specification does not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions and beneficial effects described in the present embodiment can also be combined in any appropriate manner. Those skilled in the art will understand that the embodiment can be implemented without one or more specific technical features, technical solutions or beneficial effects of a specific embodiment. In other embodiments, additional technical features and beneficial effects can also be identified in specific embodiments that do not embody all embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of a method for diagnosing a fault of a motor shaft bearing provided in an embodiment of the present application; Figure 2 A schematic diagram of a motor shaft bearing fault diagnosis device is provided in accordance with an embodiment of the present application; Figure 3 A schematic diagram of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The terms "first", "second" and "third" in this application specification and the accompanying drawings are used to distinguish different objects rather than to limit a specific order.

[0020] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0021] As new energy vehicle technology evolves, transmissions are gradually moving towards high-voltage platforms, lightweight construction, and high-speed, high-power output. These factors exacerbate the risk of motor shaft bearing failure. Coupled with the fact that most motors now use variable-frequency drives, the risk of bearing electrocorrosion increases dramatically. The different stages of electrocorrosion in transmission motor shaft bearings are as follows: 1. In the early stage, the roller surface is dark, and the electric drive system has no obvious abnormal performance; 2. As the electrical corrosion intensifies, pits appear on the inner and outer raceways of the bearings, the bearing temperature begins to rise, and the electric drive system begins to make abnormal noises; 3. As corrosion progresses, the bearing temperature rises further, the bearing lubrication performance decreases, and washboard patterns appear on the inner and outer ring raceways of the bearings. The electric drive system produces noticeable abnormal noise, and the system transmission efficiency further decreases. 4. Due to the further increase in temperature and the intensification of corrosion, the lubrication of the bearings completely fails, the rollers and the inner and outer rings of the bearings begin to deform, and eventually the bearings are completely damaged such as running rings; the motor has problems such as bore scraping and insulation; the electric drive system fails and eventually loses its function.

[0022] The essence of bearing electrical corrosion is the oil film between the bearing ball and raceway. When the insulation withstand voltage threshold formed by the insulation properties, temperature characteristics and actual thickness of the material is lower than the applied voltage between the bearings, the insulating material is broken down and electric spark discharge (EDM) is formed, forming high-temperature melting corrosion of the metal material between the ball and raceway.

[0023] In the early stages of electrocorrosion, the roller surfaces and inner and outer raceways are intact, similar to intact capacitor plates, and the capacitor's load capacity is unaffected. However, at this point, the capacitor breaks down, often due to the shaft voltage exceeding the capacitor's load capacity. As electrocorrosion progresses, pits appear on the inner and outer raceways of the bearing, reducing the electrode plate area. These pits increase the distance between the plates, reducing the capacitance and further decreasing the capacitor's load capacity, further exacerbating the corrosion process. In other words, once electrocorrosion begins in a motor shaft bearing, the corrosion progresses rapidly until failure.

[0024] It is clear that there is a need to diagnose the fault of the motor shaft bearing. However, the current fault diagnosis methods all have the problem of hysteresis.

[0025] In view of this, the present application provides a fault diagnosis method for the bearing of a motor shaft, which provides an innovative solution to the problem of lag in vibration analysis, temperature detection, and acoustic detection methods. Traditional detection methods require that the bearing fault develops to the point where the vibration is significantly aggravated, the temperature is significantly increased, or the noise is generated significantly before a judgment can be made. However, this method uses the radio frequency signal generated by the electric spark discharge of the bearing to obtain key information in the initial stage of the fault. By analyzing the pulse density, peak factor, and target frequency band energy ratio of the radio frequency signal, combined with the generated baseband signal and the duration of the preset proportional pulse, it is possible to capture early abnormalities of the bearing in real time and keenly without relying on the physical characteristics changes when the fault develops to a specific stage. This comprehensive diagnostic method based on radio frequency signal characteristics effectively overcomes the lag defects of traditional methods, realizes advanced identification and early warning of faults, greatly improves the timeliness and accuracy of diagnosis, provides reliable protection for the stable operation of the motor system, and avoids equipment damage and downtime losses caused by delayed fault discovery. It has significant technical advantages and practical value in industrial applications.

[0026] To make the technical solution of this application clearer and easier to understand, the technical solution of this application is described below in conjunction with the accompanying drawings. The fault diagnosis method for the motor shaft bearing provided in the embodiment of this application can be applied to a processing device, which can be a vehicle-mounted terminal or a device with data processing capabilities such as a processor on a vehicle. This application does not specifically limit the type of processing device.

[0027] like Figure 1 As shown in FIG, this figure is a flow chart of a method for diagnosing a fault of a motor shaft bearing provided by an embodiment of the present application. The method includes: S101: A processing device obtains a radio frequency signal generated by spark discharge of a bearing of a motor shaft, as well as a pulse density and a crest factor of the radio frequency signal, and an energy ratio of the radio frequency signal in a target frequency band.

[0028] The pulse density (CPS) of an RF signal refers to the number of RF pulse signals received per unit time (per second). Pulse density is a measure of the frequency of pulses in an RF signal. For example, if 100 RF pulses are detected within 1 second, the pulse density is 100 CPS. A higher pulse density may indicate that the signal source is emitting pulses more frequently or that a large number of pulse events occur within a specific observation period. Pulse density is a key parameter for analyzing RF signal characteristics, assessing signal strength and activity, and performing target detection and identification. By monitoring and analyzing pulse density, we can understand signal behavior patterns, thereby enabling performance evaluation, health checks, or fault diagnosis of related systems. Pulse density can help determine the persistence and development trend of a fault, avoiding the potential for missed detection by traditional methods due to weak fault signals.

[0029] In the process of acquiring the radio frequency signal, anti-interference processing can also be performed. For example, electromagnetic shielding technology, filtering circuits and other solutions can be used for anti-interference processing.

[0030] In other examples, after obtaining the RF signal, the RF signal can also be screened based on the set dynamic threshold. For example, if the signal strength of the RF signal does not meet the requirements of the dynamic threshold, this part of the data is removed from the data set, thereby ensuring the accuracy of the subsequent diagnostic results and avoiding large errors in the subsequent diagnostic process due to poor signal strength.

[0031] The peak factor (CFM) is the ratio of the peak amplitude of an RF signal to its RMS value. The CFM of an RF signal reflects the relative strength of extreme impact components within the signal. During normal bearing operation, the discharge pulse energy is low, the difference between the signal peak and RMS value is small, and the CFM is close to 1 or remains within a stable range. However, when a bearing has a severe local defect (such as sudden metal-to-metal contact discharge or transient stress concentration in a crack), extremely high-energy transient pulses are generated, significantly increasing the CFM (far greater than 1). The CFM is also used to capture transient high-energy impact signatures in the signal, identifying sudden bearing failures (such as individual balls violently impacting a defective area) or intermittent discharge events, thereby compensating for the insensitivity of traditional vibration / temperature detection to transient anomalies.

[0032] The energy ratio of the RF signal in the target frequency band refers to the ratio of the signal energy in the target frequency band to the total energy of the entire frequency band under investigation, which is used to reflect the concentration of energy in the target frequency band. Taking the target frequency band of 30-300MHz as an example, the energy ratio of the target frequency band is:

[0033] in, Represents the energy ratio of the target frequency band, It represents the power corresponding to frequency f. The numerator represents the sum of the powers corresponding to each frequency in the target frequency band, and the denominator represents the sum of the powers corresponding to each frequency from 0 to 1 GHz.

[0034] In some embodiments, the processing device can obtain a first frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a Rogowski coil array, and obtain a second frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a log-periodic antenna; calculate the average frequency of the first frequency and the second frequency; if the average frequency is lower than a preset frequency, use the Rogowski coil array to obtain the radio frequency signal generated by the spark discharge of the bearing of the motor shaft; if the average frequency is greater than the preset frequency, use the log-periodic antenna to obtain the radio frequency signal generated by the spark discharge of the bearing of the motor shaft; if the average frequency is equal to the preset frequency, continue to obtain the first frequency and the second frequency respectively until the average frequency is not equal to the preset frequency.

[0035] In an embodiment of the present application, the processing device collaboratively acquires the radio frequency signal of the bearing electric spark discharge through the Rogowski coil array and the logarithmic periodic antenna, and dynamically selects the sensor based on the average frequency of the first frequency and the second frequency, thereby achieving adaptive matching of the signal frequency characteristics: in the low frequency band, the high sensitivity of the Rogowski coil array is used to accurately capture weak signals, and in the high frequency band, the wideband characteristics of the logarithmic periodic antenna are used to fully collect the radiation signal, effectively solving the performance limitations of a single sensor in different frequency bands, improving the accuracy and completeness of full-band signal acquisition, providing reliable data support for the early identification and accurate diagnosis of bearing faults, and significantly enhancing the system's adaptability and anti-interference ability under complex working conditions.

[0036] S102: The processing device generates a baseband signal of a target frequency band according to the radio frequency signal, and determines a duration of a pulse of a preset proportion according to the radio frequency signal.

[0037] In order to enhance anti-interference capability and reduce signal processing complexity, the processing equipment can generate a baseband signal of the target frequency band based on the RF signal. Converting the RF signal into a baseband signal can remove the high-frequency carrier through frequency shifting, reducing signal processing complexity and hardware cost, facilitating direct extraction of original features (such as pulses, frequency, and energy distribution), enhancing anti-interference capability and adapting to the digital signal processing architecture, providing an efficient and reliable low-frequency signal foundation for subsequent fault diagnosis.

[0038] In some embodiments, the processing device may first adjust the frequency of the local oscillator to the center frequency of the target frequency band, with the goal of moving the RF signal in the target frequency band to the baseband through a mixing operation. For example, if the target frequency band is 30-300 MHz, the center frequency is 165 MHz, and the local oscillator frequency is set to 165 MHz for subsequent processing. The processing device then determines the in-phase component and the quadrature component of the orthogonal local oscillator signal based on the center frequency and the RF signal;

[0039]

[0040] in, represents the in-phase component of the orthogonal local oscillator signal, represents the orthogonal component of the orthogonal local oscillator signal, represents the frequency of the local oscillator, Indicates a radio frequency signal.

[0041] By multiplying the in-phase and quadrature components of the quadrature local oscillator signal with the RF signal, and then square-summing the resulting sum, the baseband signal in the target frequency band is obtained. This processing method effectively extracts useful information from the RF signal while suppressing noise and interference.

[0042] In some embodiments, the processing device determines the duration of a preset proportion of pulses based on the radio frequency signal. The specific process is as follows: first, the radio frequency signal is subjected to time-frequency analysis (such as through STFT short-time Fourier transform), decomposed to obtain a time-frequency matrix, extract each pulse waveform and perform interpolation processing to improve accuracy; then, the time difference corresponding to the amplitude from 10% to 90% of the amplitude of each pulse waveform is calculated, and this time difference is the duration of a single pulse; finally, according to a preset proportion (such as selecting a certain proportion of pulses by energy or number, such as the top 20% of pulses in energy), the corresponding pulses are screened out, and the duration of these pulses is determined for subsequent analysis. For example, a time difference of less than 10ns indicates an electro-corrosion feature, which provides a transient feature basis for bearing fault diagnosis.

[0043] S103: The processing device diagnoses the fault of the motor shaft bearing according to the pulse density, crest factor, energy ratio, baseband signal and duration.

[0044] Among them, the pulse density can reflect the frequency of discharge events. The higher the density, the more active the bearing surface defects and the more serious the fault may be. If the peak factor is large, it means that there are strong impact components in the signal, which may correspond to serious local defects in the bearing (such as sudden violent discharge or mechanical impact). If the energy ratio is higher, it means that the discharge energy is more concentrated in the target frequency band (30-300MHz), indicating that there is abnormal energy distribution related to the fault in this frequency band. Baseband signal: directly retains the original characteristics of the RF signal after down-conversion, which is used to observe the signal time domain and frequency domain details and assist in judging the fault type and characteristics. Duration: calculate the time difference corresponding to the 10%-90% amplitude of the pulse waveform. If the time difference is less than the difference threshold, it is marked as an electro-corrosion feature, indicating that the bearing has an electro-corrosion fault.

[0045] The processing equipment integrates these parameters: when the pulse density is high (frequent discharges), the crest factor exceeds the threshold (strong impact), the energy ratio is abnormal (energy concentration in the target frequency band), and the duration matches the characteristics of electrocorrosion, it can be determined that the bearing has electrocorrosion or a similar fault. If a single parameter is abnormal, further analysis is combined with other parameters. For example, a high energy ratio alone with a low pulse density may indicate early-stage partial discharge. Through multi-dimensional parameter cross-validation, accurate identification and diagnosis of motor shaft bearing faults are achieved.

[0046] In an embodiment of the present application, the fault of the motor shaft bearing is diagnosed based on the pulse density, crest factor, energy ratio, baseband signal and duration, including: The pulse density, peak factor, energy ratio, baseband signal and duration are vectorized to obtain a multi-dimensional RF vector; the multi-dimensional RF vector is input into the fault diagnosis model of the motor bearing to obtain a fault diagnosis result; wherein, the fault diagnosis model of the motor bearing is trained based on sample data, and the sample data includes sample features and sample labels, and the sample features include sample pulse density, sample peak factor, sample energy ratio, sample baseband signal and sample duration.

[0047] In an embodiment of the present application, a method for diagnosing motor bearing faults using multidimensional features includes: first, vectorizing key features such as pulse density, crest factor, energy ratio, baseband signal, and duration to construct a multidimensional RF vector. These vectors are then input into a fault diagnosis model pre-trained based on sample data to obtain a fault diagnosis result. The sample data includes sample features (such as sample pulse density) and sample labels, which are used to train the model to learn the mapping between fault features and fault types.

[0048] Multi-dimensional RF vectors integrate multiple key features, providing a more comprehensive and accurate reflection of the motor bearing's operating status than a single feature alone. Different fault types may manifest uniquely in different features, and the combination of multiple features can capture even more subtle fault differences, reducing the likelihood of misdiagnosis and missed diagnoses.

[0049] For example, pulse density alone may not be able to distinguish different types of wear faults, but combined with features such as peak factor and energy ratio, the fault type can be determined more accurately.

[0050] In some examples, sample data is first prepared. This involves collecting historical data containing sample features (pulse density, crest factor, energy ratio, baseband signal, and duration) and sample labels (fault type / fault presence) for training the fault diagnosis model. Next, the features are vectorized. This involves converting real-time acquired parameters such as pulse density, crest factor, energy ratio, baseband signal, and duration into multi-dimensional RF vectors (digitized feature vectors) to serve as model input. Finally, the model is trained based on the sample data to learn the mapping between features and fault labels. The principles of the diagnostic process are detailed below.

[0051] Each dimension in the RF vector (such as pulse density and crest factor) corresponds to a signal characteristic of the motor bearing during operation. These characteristics are indirect representations of the fault state. The specific relationship is as follows: Pulse density: reflects the density of abnormal pulses in the signal. Faults such as bearing wear and cracks will lead to an increase in impact signals and higher pulse density. Peak factor: measures the ratio of the signal peak to the effective value. Sudden faults (such as spalling) will significantly increase the peak factor. Energy ratio: the energy distribution of different frequency components. Bearing faults will cause abnormal energy proportions of specific frequencies (such as fault characteristic frequencies). Baseband signal: the low-frequency component of the original signal, which directly reflects the basic characteristics of mechanical vibration. Faults will change its waveform shape (such as periodic shocks). Duration: the duration of the abnormal component in the signal, which is related to the stage of fault development (such as short duration of early faults and prolonged duration in later stages). Different fault types will result in different combination patterns of the above features. The multidimensional vector is equivalent to "encoding" the fault as a point in high-dimensional space. Points of the same type of faults are clustered in space, while points of different types of faults are separated from each other.

[0052] The essence of fault diagnosis models (such as machine learning and deep learning models) is a nonlinear classifier in high-dimensional space, and its working principle can be divided into two steps.

[0053] Step 1: Learn the mapping relationship between features and faults.

[0054] The input is a multidimensional RF vector (feature) of historical samples and corresponding fault labels (e.g., "normal," "inner race wear," "rolling element failure"). Through manually designed feature engineering (e.g., statistics, frequency domain metrics) or automated feature selection (e.g., PCA dimensionality reduction), the original features are mapped to a more discriminative space. Alternatively, a hierarchical representation of features is automatically learned through the hidden layers of a neural network (e.g., extracting abstract features such as "local impact, failure mode, global category" layer by layer from the original signal). Using a loss function (e.g., cross-entropy) and an optimization algorithm (e.g., stochastic gradient descent), the model parameters are adjusted to ensure that vectors of similar faults in the feature space are as close as possible and vectors of heterogeneous faults are as far apart as possible.

[0055] Step 2: Real-time judgment.

[0056] When a multi-dimensional RF vector collected in real time is input into the model, the model will map the input vector to a decision space (such as classification probability) using trained parameters, determine which fault category the vector belongs to based on a preset threshold or maximum probability value, and then output the fault diagnosis result.

[0057] In some embodiments, if the fault diagnosis result of the motor shaft bearing by the processing device indicates that there is a fault in the motor shaft bearing, the fault type of the motor shaft bearing is determined. At the same time, the bearing fault can also be displayed through the vehicle system to prompt the user.

[0058] In some embodiments, after determining that a bearing of the motor shaft is faulty, the processing device may further obtain the bearing temperature of the motor shaft and the shaft voltage at both ends of the bearing of the motor shaft; if the bearing temperature of the motor shaft is greater than a preset temperature, the fault type is determined to be lubrication failure; or, if the shaft voltage at both ends of the bearing of the motor shaft is greater than a preset voltage, the fault type is determined to be insulation failure of the motor shaft. After determining that the motor shaft bearing is faulty, the processing equipment further diagnoses by obtaining the bearing temperature and the shaft voltage at both ends. If the bearing temperature is greater than the preset temperature, it is determined to be a lubrication failure (abnormal frictional heat generation due to poor lubrication). If the shaft voltage at both ends of the bearing is greater than the preset voltage, it is determined to be a failure of the motor shaft insulation layer (damage to the insulation layer causes current leakage to form abnormal shaft voltage). This method uses key physical parameters such as temperature and shaft voltage to accurately distinguish the type of fault, providing a basis for targeted maintenance (such as replenishing lubricating oil or inspecting the insulation layer), avoiding blind repairs, reducing maintenance costs, and providing early warning of faults to prevent the fault from worsening, effectively improving the reliability and stability of the motor system operation.

[0059] In some embodiments, the processing device can also determine whether the shaft voltage at both ends of the bearing of the motor shaft meets the characteristics of capacitive charging and discharging. For example, the processing device can determine the similarity between the voltage characteristic diagram of the shaft voltage and the capacitive charging and discharging characteristic diagram. If the similarity is greater than the similarity threshold, it is considered that the shaft voltage meets the characteristics of capacitive charging and discharging. Otherwise, it does not meet the similarity threshold. The similarity threshold can be 90%.

[0060] Based on the above description, the embodiment of the present application provides a method for diagnosing bearing faults of motor shafts, which effectively solves the problem of lag in vibration analysis, temperature detection, and acoustic detection methods. Traditional methods rely on the development of faults to a certain stage (such as intensified vibration, significantly increased temperature, and significant noise) before they can be detected. However, this method uses the radio frequency signal generated by the electric spark discharge of the bearing to capture key information at the early stage of the fault. By obtaining the pulse density, peak factor, and target frequency band energy ratio of the radio frequency signal, combined with the generated baseband signal and pulse duration for comprehensive analysis, the bearing status can be reflected in real time and sensitively. This multi-dimensional diagnostic method based on radio frequency signal characteristics breaks through the traditional method's dependence on the degree of fault development, realizes advanced identification of early faults, greatly improves the timeliness and accuracy of diagnosis, provides more efficient early warning for motor system maintenance, reduces the risk of downtime due to delayed fault discovery, and ensures stable equipment operation. It has significant technical advantages and practical value.

[0061] Combined with the above Figure 1 The fault diagnosis method of the motor shaft bearing provided by the embodiment of the present application is described in detail. Figure 2 The fault diagnosis device for the bearing of the motor shaft provided in an embodiment of the present application is introduced.

[0062] like Figure 2 As shown in FIG, this figure is a schematic diagram of a fault diagnosis device for a motor shaft bearing provided by an embodiment of the present application, the device comprising: An acquisition module 201 is configured to acquire a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft, as well as a pulse density, a crest factor, and an energy ratio of the radio frequency signal in a target frequency band. a determination module 202 configured to generate a baseband signal of a target frequency band based on the radio frequency signal, and determine a duration of pulses of a preset proportion based on the radio frequency signal; The diagnosis module 203 is configured to diagnose a fault of the bearing of the motor shaft according to the pulse density, the peak factor, the energy ratio, the baseband signal, and the duration.

[0063] Optionally, the diagnostic module 203 is specifically used to vectorize the pulse density, the peak factor, the energy ratio, the baseband signal and the duration to obtain a multi-dimensional RF vector; input the multi-dimensional RF vector into the fault diagnosis model of the motor bearing to obtain a fault diagnosis result; wherein, the fault diagnosis model of the motor bearing is trained based on sample data, the sample data includes sample features and sample labels, and the sample features include sample pulse density, sample peak factor, sample energy ratio, sample baseband signal and sample duration.

[0064] Optionally, the diagnosis module 203 is further configured to determine a fault type of the motor shaft bearing if a fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty.

[0065] Optionally, the acquisition module 201 is also used to obtain the bearing temperature of the motor shaft and the shaft voltage at both ends of the bearing of the motor shaft; the diagnosis module 203 is also used to determine that the fault type is lubrication failure if the bearing temperature of the motor shaft is greater than a preset temperature; or, if the shaft voltage at both ends of the bearing of the motor shaft is greater than a preset voltage, determine that the fault type is failure of the insulation layer of the motor shaft.

[0066] Optionally, the acquisition module 201 is specifically used to acquire a first frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a Rogowski coil array, and acquire a second frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a log-periodic antenna; calculate the average frequency of the first frequency and the second frequency; if the average frequency is lower than a preset frequency, acquire the radio frequency signal generated by the spark discharge of the bearing of the motor shaft by using the Rogowski coil array; if the average frequency is greater than the preset frequency, acquire the radio frequency signal generated by the spark discharge of the bearing of the motor shaft by using the log-periodic antenna.

[0067] Optionally, the determination module 202 is specifically used to adjust the frequency of the local oscillator to the center frequency of the target frequency band; determine the in-phase component and the quadrature component of the orthogonal local oscillator signal based on the center frequency and the radio frequency signal; and determine the sum of the squares of the in-phase component and the quadrature component of the orthogonal local oscillator signal as the baseband signal of the target frequency band.

[0068] Optionally, the diagnosis module 203 is further configured to give a prompt through the vehicle computer system if the fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty.

[0069] The fault diagnosis device for the motor shaft bearing according to the embodiment of the present application may correspond to the method described in the embodiment of the present application, and the above-mentioned other operations and / or functions of each module / unit of the fault diagnosis device for the motor shaft bearing are respectively for realizing Figure 1 For the sake of brevity, the corresponding processes of the various methods in the illustrated embodiments are not described again here.

[0070] The present application also provides a computing device. Figure 3 As shown, this figure is a schematic diagram of a computing device provided by an embodiment of the present application, and the computing device 300 includes a bus 301, a processor 302, a communication interface 303 and a memory 304. The processor 302, the memory 304 and the communication interface 303 communicate with each other via the bus 301.

[0071] The bus 301 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0072] The processor 302 may be any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0073] The communication interface 303 is used for communicating with the outside.

[0074] The memory 304 may include volatile memory, such as random access memory (RAM). The memory 304 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0075] The memory 304 stores executable codes, and the processor 302 executes the executable codes to perform the aforementioned motor shaft bearing fault diagnosis method.

[0076] Specifically, in the implementation Figure 2 In the case of the embodiment shown, and Figure 2 When each module or unit of the motor shaft bearing fault diagnosis device described in the embodiment is implemented by software, the execution Figure 2 The software or program code required for the functions of each module / unit in the system may be partially or completely stored in the memory 304. The processor 302 executes the program code corresponding to each unit stored in the memory 304 to perform the aforementioned fault diagnosis method for the motor shaft bearing.

[0077] Embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium can be any available medium capable of being stored by a computing device, or a data storage device such as a data center that contains one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive). The computer-readable storage medium includes instructions that instruct the computing device to execute the aforementioned motor shaft bearing fault diagnosis method.

[0078] The present application also provides a computer program product comprising one or more computer instructions that, when loaded and executed on a computing device, fully or partially generate the process or function described in the present application.

[0079] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0080] When the computer program product is executed by a computer, the computer performs any of the aforementioned methods for diagnosing a fault in a motor shaft bearing. The computer program product may be a software installation package. When any of the aforementioned methods for diagnosing a fault in a motor shaft bearing is required, the computer program product may be downloaded and executed on the computer.

[0081] The descriptions of the processes or structures corresponding to the above figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0082] The above description is only a specific implementation method of the present application, but the protection scope of the present application is not limited thereto. Any changes or replacements within the technical scope disclosed in the present application should be included in the protection scope of the present application.

Claims

1. A method for diagnosing a fault of a motor shaft bearing, characterized in that: The method comprises: Acquire a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft, as well as a pulse density, a crest factor, and an energy ratio of the radio frequency signal in a target frequency band; generating a baseband signal of a target frequency band according to the radio frequency signal, and determining a duration of a pulse of a preset proportion according to the radio frequency signal; A fault of a bearing of the motor shaft is diagnosed according to the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration.

2. The method according to claim 1, characterized in that The diagnosing a fault of a bearing of the motor shaft according to the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration includes: Vectorizing the pulse density, the crest factor, the energy ratio, the baseband signal, and the duration to obtain a multi-dimensional radio frequency vector; Inputting the multi-dimensional radio frequency vector into a fault diagnosis model of a motor bearing to obtain a fault diagnosis result; The motor bearing fault diagnosis model is trained based on sample data, wherein the sample data includes sample features and sample labels, and the sample features include sample pulse density, sample peak factor, sample energy ratio, sample baseband signal and sample duration.

3. The method according to claim 1, characterized in that The method further comprises: If the fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty, the fault type of the motor shaft bearing is determined.

4. The method according to claim 3, characterized in that Determining the fault type of the bearing of the motor shaft includes: Acquiring the bearing temperature of the motor shaft and the shaft voltage at both ends of the bearing of the motor shaft; If the bearing temperature of the motor shaft is greater than a preset temperature, the fault type is determined to be lubrication failure; or if the shaft voltage across the bearing of the motor shaft is greater than a preset voltage, the fault type is determined to be insulation layer failure of the motor shaft.

5. The method according to claim 1, wherein The obtaining of the radio frequency signal generated by the electric spark discharge of the bearing of the motor shaft includes: Acquire a first frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a Rogowski coil array, and acquire a second frequency of the radio frequency signal generated by the spark discharge of the bearing of the motor shaft through a logarithmic periodic antenna; calculating an average frequency of the first frequency and the second frequency; If the average frequency is lower than a preset frequency, a Rogowski coil array is used to obtain a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft; If the average frequency is greater than the preset frequency, a log-periodic antenna is used to acquire a radio frequency signal generated by the spark discharge of the bearing of the motor shaft.

6. The method according to claim 1, wherein Generating a baseband signal of a target frequency band according to the radio frequency signal includes: Adjust the frequency of the local oscillator to the center frequency of the target frequency band; Determining an in-phase component and a quadrature component of an orthogonal local oscillator signal according to the center frequency and the radio frequency signal; The sum of the squares of the in-phase component and the quadrature component of the quadrature local oscillator signal is determined as the baseband signal of the target frequency band.

7. The method according to claim 1, characterized in that The method further comprises: If the fault diagnosis result of the motor shaft bearing indicates that the motor shaft bearing is faulty, a prompt is given through the vehicle computer system.

8. A fault diagnosis device for a motor shaft bearing, characterized in that: The device comprises: an acquisition module, configured to acquire a radio frequency signal generated by an electric spark discharge of a bearing of the motor shaft, as well as a pulse density, a crest factor, and an energy ratio of the radio frequency signal in a target frequency band; a determination module, configured to generate a baseband signal of a target frequency band based on the radio frequency signal, and determine a duration of pulses of a preset proportion based on the radio frequency signal; A diagnostic module is used to diagnose a fault of the bearing of the motor shaft according to the pulse density, the peak factor, the energy ratio, the baseband signal and the duration.

9. A computing device, characterized in that including memory and processor; One or more computer programs are stored in the memory, and the one or more computer programs include instructions; when the instructions are executed by the processor, the computing device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store a computer program, and the computer program is used to execute the method according to any one of claims 1 to 7.

Images