A method and system for identifying the air-gap eccentricity fault of a permanent magnet generator

Abstract

The present invention discloses a method and system for identifying the air-gap eccentricity fault of a permanent magnet generator, which relates to the technical field of permanent magnet generators. By using wavelet transform to extract time-frequency features and generate multi-band feature vectors, the system can more sensitively identify the subtle features of the air-gap thickness deviation, effectively improving the extraction and analysis accuracy of feature signals in different frequency bands. Secondly, based on the feature vectors, the situation of the air-gap thickness deviation is analyzed, an eccentricity degree index is constructed, and the eccentricity degree is quantified, which can intuitively judge the severity of the air-gap eccentricity fault of the permanent magnet generator. When the eccentricity degree exceeds the set threshold, the system automatically issues a fault type troubleshooting instruction, making the fault detection process more intelligent. Subsequently, through infrared thermal data and temperature gradient analysis of the abnormal changes at each position, the air-gap foreign object fault that may cause air-gap eccentricity is further identified, making up for the limitations of single-signal analysis in traditional detection.

Description

Technical Field

[0001] The present invention relates to the technical field of permanent magnet generators, and specifically to a method and system for identifying the air-gap eccentricity fault of a permanent magnet generator. Background Art

[0002] As a key device in new energy technology, permanent magnet generators are widely used in scenarios such as wind power generation, electric vehicles, and industrial automation due to their high efficiency and stability. However, during the operation of permanent magnet generators, with the aging of mechanical components, the influence of environmental factors, and the change of loads, they often face problems such as air-gap eccentricity faults. The air-gap eccentricity fault is mainly manifested as the uneven air gap between the rotor and the stator, resulting in the asymmetry of the motor magnetic field distribution, triggering abnormal phenomena such as vibration and heating, and further affecting the overall performance and service life of the generator. Therefore, the identification of the air-gap eccentricity fault of permanent magnet generators has important practical significance.

[0003] For the identification of the air-gap eccentricity fault of permanent magnet generators, current traditional methods mostly rely on vibration analysis or single magnetic field signal analysis. However, these methods have certain limitations in detection accuracy and accuracy. Since the air-gap eccentricity fault of permanent magnet generators has the characteristics of multi-frequency bands and the signals are often non-stationary, single analysis will cause the system to lack the support for comprehensive analysis of multi-source signals. In addition, traditional detection methods usually ignore the local thermal anomalies that may be caused by foreign objects in the air gap and lack the effective utilization of temperature characteristics, thus reducing the comprehensive diagnostic ability for air-gap eccentricity faults. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a method and system for identifying the air-gap eccentricity fault of a permanent magnet generator, which solves the problems in the above background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for identifying the air-gap eccentricity fault of a permanent magnet generator includes the following steps.

[0006] S1. Deploy a number of groups of sensors inside the permanent magnet generator in advance, and based on the number of groups of sensors, real-time monitor the vibration signal and magnetic field signal of the permanent magnet generator during operation to construct the motor operation data information.

[0007] S2. Perform data preprocessing on the motor operation data information, and use wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal to generate the feature vector during motor operation.

[0008] S3. Based on the feature vector formed after data preprocessing in S2, preliminarily analyze whether there is a deviation in the air-gap thickness of the permanent magnet generator to construct the eccentricity degree index Pczb. Based on the value of the eccentricity degree index Pczb, determine whether to send out a fault type troubleshooting instruction outward.

[0009] S4. After receiving the fault type troubleshooting instruction, use thermal imaging technology to monitor the two-dimensional temperature distribution in the air gap area of the permanent magnet generator to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator, and construct the temperature gradient at each position. ;

[0010] S5. Preset a safety threshold Q, and by comparing and analyzing the temperature gradient at each position with the safety threshold Q, determine the fault type of the current permanent magnet generator.

[0011] Preferably, the specific steps of S1 include:

[0012] S11. Deploy several groups of sensors inside the permanent magnet generator in advance to form a sensor group, and the sensor group includes vibration sensors and Hall sensors;

[0013] S12. According to the sensor group formed in S11, monitor and obtain the vibration signal and magnetic field signal of the permanent magnet generator during operation in real time to construct the motor operation data information.

[0014] Preferably, the specific steps of S2 include:

[0015] S21. Remove the noise in the motor operation data information through wavelet denoising technology, and remove the trend component in the motor operation data information through the moving average method to perform data preprocessing on the motor operation data information. After data preprocessing, use wavelet transform technology to extract the time-frequency characteristics of the motor operation data information to generate the feature vector during motor operation, and the feature vector during motor operation includes the magnetic flux density value in the air gap and vibration amplitude .

[0016] Preferably, the specific steps of S3 include:

[0017] S31. According to the feature vector during motor operation, after feature extraction, analyze the change of the magnetic flux density in the air gap of the permanent magnet generator over time to calculate and obtain the magnetic field strength change rate in different frequency bands , and the magnetic field strength change rate in different frequency bands is obtained through the following formula:

[0018] ;

[0019] In the formula, represents the magnetic flux density value in the air gap, represents the time interval, is the derivative of the magnetic flux density with respect to time;

[0020] S32. After feature extraction based on the eigenvector during the operation of the motor, analyze the synchronization between the vibration signal and the magnetic field signal of the permanent magnet generator to calculate and obtain the phase difference between the vibration signal and the magnetic field signal in different frequency bands. , which is specifically obtained according to the following formula:

[0021] ;

[0022] In the formula, represents the vibration amplitude, represents the vibration amplitude and the dot product of the magnetic flux density value in the air gap is used to calculate the phase relationship between the two. The dot product is a scalar of the interaction between two signals, reflecting the amplitudes and relative directions of the two signals; represents the modulus of the vibration amplitude , represents the modulus of the magnetic flux density value in the air gap, is the inverse cosine function.

[0023] Preferably, the specific steps of S3 further include:

[0024] S33. Based on the magnetic field strength change rate obtained in S31 and S32 and the phase difference

[0025] between the vibration signal and the magnetic field signal in different frequency bands, preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator, and after linear normalization processing, construct an eccentricity index Pczb. The eccentricity index Pczb is obtained through the following formula:

[0026] In the formula, represents the number of frequency bands, , represents the th magnetic field strength change rate in the frequency band, represents the average magnetic field strength change rate, represents the th phase difference between the vibration signal and the magnetic field signal in the frequency band, represents the average value of the phase difference between the vibration signal and the magnetic field signal, and are both weight values, represents the first correction constant, where and The specific values are set by the user according to the situation.

[0027] Preferably, the specific steps of S3 further include:

[0028] S34. Preset a judgment threshold P, and compare and analyze it with the eccentricity degree index Pczb to determine whether there is a deviation in the air gap thickness of the current permanent magnet generator. The specific analysis content is as follows:

[0029] If the eccentricity degree index Pczb exceeds the judgment threshold P, it is determined that there is a deviation in the air gap thickness of the current permanent magnet generator, and at this time, a fault type troubleshooting instruction will be sent outwards;

[0030] If the eccentricity degree index Pczb does not exceed the judgment threshold P, it is determined that there is no deviation in the air gap thickness of the current permanent magnet generator for the time being, and at this time, a green warning notice will be sent outwards.

[0031] Preferably, the specific steps of S4 include:

[0032] S41. After receiving the fault type troubleshooting instruction issued in S34, use a thermal imaging device to capture the infrared thermal image of the air gap area in the permanent magnet generator, and use Gaussian filtering technology to denoise the infrared thermal image. Among them, the thermal imaging device refers to an infrared thermal imager;

[0033] S42. According to the infrared thermal image after image denoising, obtain infrared thermal data, and the infrared thermal data includes the temperature values at each pixel position in the infrared thermal image ;

[0034] S43. By comparing the temperature values at each pixel position in the infrared thermal image with a preset threshold, if the temperature values at each pixel position in the infrared thermal image exceed the preset threshold, the corresponding pixel position will be marked as an abnormal point, and by counting the abnormal points, the number of abnormal points YS can be obtained.

[0035] Preferably, the specific steps of S4 further include:

[0036] S43. According to the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator to construct the temperature gradient at each pixel position , and the specific method for obtaining it is as follows:

[0037] ;

[0038] In the formula, represents the temperature gradient along the horizontal axis direction; represents the temperature gradient along the vertical axis direction; x and y respectively represent the horizontal and vertical coordinates in the infrared thermal image.

[0039] Preferably, the specific steps of S5 include:

[0040] S51. Compare and analyze the temperature gradient at each position with the safety threshold Q to determine the fault type of the current permanent magnet generator. The specific judgment content is as follows:

[0041] S511. If the temperature gradient at the corresponding position exceeds the safety threshold Q, mark the current position as a local hot spot at this time;

[0042] S512. If the temperature gradient at the corresponding position does not exceed the safety threshold Q, do not mark the current position as a local hot spot for the time being;

[0043] S52. Count the local hot spots obtained in S511 to obtain the number of local hot spots JS. When the number of local hot spots JS exceeds 80% of the number of abnormal points YS, it is determined at this time that there is a foreign object fault in the air gap of the current permanent magnet generator.

[0044] A permanent magnet generator air gap eccentricity fault identification system includes an acquisition module, a processing module, an eccentricity analysis module, a type troubleshooting module, and a type feedback module;

[0045] The acquisition module is used to deploy several groups of sensors inside the permanent magnet generator in advance, and based on the several groups of sensors, it monitors the vibration signal and magnetic field signal of the permanent magnet generator during operation in real time to construct motor operation data information;

[0046] The processing module is used to preprocess the motor operation data information and use wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal to generate a feature vector during motor operation;

[0047] The eccentricity analysis module is used to preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator based on the feature vector formed after data preprocessing, to construct an eccentricity degree index Pczb, and based on the value of the eccentricity degree index Pczb, determine whether to send out a fault type troubleshooting instruction externally;

[0048] The type troubleshooting module is used to, when receiving the fault type troubleshooting instruction, use thermal imaging technology to monitor the two-dimensional temperature distribution of the air gap area inside the permanent magnet generator to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of foreign objects in the air gap area inside the permanent magnet generator, and construct the temperature gradient at each position ;

[0049] The type feedback module is used to preset a safety threshold Q, by comparing the temperature gradient at each position Compare and analyze with the safety threshold Q to determine the fault type of the current permanent magnet generator.

[0050] The present invention provides a method and system for identifying the air-gap eccentricity fault of a permanent magnet generator, which has the following beneficial effects:

[0051] (1) Through the deployment of the sensor group in step S1, the method can monitor and obtain vibration signals and magnetic field signals in real time, enabling the fault identification method to timely capture the operation data of the permanent magnet generator under different operating states and providing multi-faceted basic data for subsequent analysis. By using wavelet transform to extract time-frequency features and generate multi-band feature vectors, the system can more sensitively identify the subtle features of the air-gap thickness deviation, effectively improving the extraction and analysis accuracy of feature signals in different frequency bands. Secondly, based on the analysis of the air-gap thickness deviation using the feature vector, an eccentricity degree index Pczb is constructed, and the eccentricity degree is quantified, which can intuitively judge the severity of the air-gap eccentricity fault of the permanent magnet generator. When the eccentricity degree exceeds the set threshold, the system automatically issues a fault type troubleshooting instruction, making the fault detection process more intelligent. Subsequently, the temperature distribution in the air-gap area is monitored by thermal imaging technology, and the abnormal changes at each position are analyzed in combination with infrared thermal data and temperature gradient, further identifying the air-gap foreign object fault that may cause air-gap eccentricity and making up for the limitations of single-signal analysis in traditional detection. Finally, in step S5, by further comparing the temperature gradient at each position with the safety threshold to count the local hot spot positions, the accurate judgment of the air-gap foreign object risk is realized. In short, this method combines multi-source data such as vibration signals, magnetic field signals, and temperature signals, not only improving the accuracy of air-gap eccentricity fault identification, but also making the troubleshooting of complex faults such as air-gap foreign objects more comprehensive, providing an effective guarantee for the safe and stable operation of the permanent magnet generator.

[0052] (2) In step S2, wavelet denoising technology is used to remove the noise from the collected motor operation data information, and the trend component in the data is removed by the moving average method, making the data fluctuations more stable and the features more prominent. This data preprocessing method improves the representativeness of the signal features and lays a more stable foundation for fault analysis. On this basis, wavelet transform technology is further used to extract time-frequency features and generate a feature vector containing the air-gap magnetic flux density value and vibration amplitude, enabling the system to more sensitively identify the local features and spectral changes of the fault signal. These processing steps not only improve the accuracy of air-gap eccentricity fault detection, but also effectively eliminate the interference components in the signal, providing high-quality feature data for subsequent analysis.

[0053] (3) First, in step S3, by analyzing the change of the magnetic flux density in the air gap in the eigenvector over time, the change rate of the magnetic field intensity in different frequency bands can be obtained. The calculation of this index accurately reflects the change characteristics of the air gap thickness, enabling the system to monitor the magnetic field intensity fluctuations of the permanent magnet generator in different frequency ranges in real time, and thus more sensitively detect the fault characteristics caused by air gap eccentricity. This method uses Hall sensors or magnetic flux density sensors to measure the time change of the magnetic field intensity, effectively improving the timeliness of fault detection and the accuracy of data. Secondly, through the analysis of the synchronization between the vibration signal and the magnetic field signal in step S32, the phase difference between the two signals in different frequency bands is further calculated, revealing the relative changes of vibration and magnetic field, thereby helping to identify the type of air gap eccentricity fault. This multi-frequency band phase difference analysis method not only increases the richness of the interaction between detection signals, but also enables the identification of fault types, making the system more accurate and comprehensive in identifying fault characteristics under complex working conditions, and providing key data information support for subsequent fault judgment and maintenance decision-making.

[0054] (4) Through the temperature gradient analysis in step S43, the local temperature change rate of each pixel point is calculated, and the system can accurately identify the position of temperature mutation in the air gap area. The analysis of temperature gradient can reveal the distribution of local hot spots and temperature anomaly points, effectively identifying the possibility of the existence of foreign objects in the air gap. This risk analysis method based on temperature gradient can not only accurately locate the fault position, but also provide an in-depth assessment of the local thermal anomaly caused by foreign objects in the air gap, providing reliable data support and diagnostic basis for further investigation of air gap eccentricity faults. Description of the Drawings

[0055] Figure 1 It is a schematic flow chart of a method for identifying air gap eccentricity faults of a permanent magnet generator according to the present invention;

[0056] Figure 2 It is a block diagram of a system for identifying air gap eccentricity faults of a permanent magnet generator according to the present invention. Detailed Embodiments

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0058] Embodiment 1

[0059] Please refer to Figure 1 , the present invention provides a method for identifying air gap eccentricity faults of a permanent magnet generator, including the following steps,

[0060] S1. Deploy several groups of sensors inside the permanent magnet generator in advance, and based on the several groups of sensors, monitor the vibration signal and magnetic field signal of the permanent magnet generator in real time during operation to construct the motor operation data information;

[0061] S2. Perform data preprocessing on the motor operation data information, and use wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal to generate the feature vector during motor operation;

[0062] S3. Based on the feature vector formed after data preprocessing in S2, preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator to construct the eccentricity degree index Pczb. Based on the value of the eccentricity degree index Pczb, determine whether to send out a fault type troubleshooting instruction;

[0063] S4. When receiving the fault type troubleshooting instruction, use thermal imaging technology to monitor the two-dimensional temperature distribution in the air gap area of the permanent magnet generator to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator to construct the temperature gradient at each position ;

[0064] S5. Preset a safety threshold Q, and by comparing and analyzing the temperature gradient at each position with the safety threshold Q, determine the fault type of the current permanent magnet generator.

[0065] In this embodiment, the method further improves the recognition accuracy and comprehensiveness of air-gap eccentricity faults through the comprehensive monitoring and analysis of multiple signals. Specifically, in step S1 of the method, by deploying multiple sensors inside the permanent magnet generator, the real-time acquisition of vibration signals and magnetic field signals is realized, ensuring the real-time performance and reliability of the data. In step S2, wavelet transform is used to perform data preprocessing and time-frequency feature extraction on the vibration signals and magnetic field signals. It not only effectively removes noise interference but also retains the key feature information of the signals, making the generated feature vectors more accurately reflect the operating state of the motor. In addition, in step S3, by constructing an eccentricity degree index, it is possible to quickly determine whether there is a deviation in the air-gap thickness. If an abnormality is detected, a fault type troubleshooting instruction is immediately triggered, further improving the response speed of the fault. In step S4, thermal imaging technology is used to perform infrared thermal imaging monitoring on the air-gap area, capturing high-precision temperature distribution information and calculating the temperature gradient at each position, further improving the detection sensitivity of the local heating phenomenon caused by foreign objects in the air-gap. Finally, in step S5, the temperature gradient is compared with the safety threshold to achieve accurate discrimination of different fault types, especially with higher accuracy and reliability in identifying the eccentricity fault caused by foreign objects in the air-gap. Therefore, the method has beneficial effects such as high real-time performance, accurate detection, and strong adaptability, providing a strong guarantee for the safe and efficient operation of the permanent magnet generator.

[0066] Embodiment 2

[0067] Please refer to Figure 1 , specifically: The specific steps of S1 include:

[0068] S11. Deploy several groups of sensors inside the permanent magnet generator in advance to form a sensor group, and the sensor group includes vibration sensors and Hall sensors;

[0069] S12. According to the sensor group formed in S11, the vibration signals and magnetic field signals of the permanent magnet generator during operation are monitored in real time to construct the motor operation data information.

[0070] The specific steps of S2 include:

[0071] S21. Remove the noise in the motor operation data information through wavelet denoising technology, and remove the trend component in the motor operation data information through the moving average method to ensure data stability and further improve the representativeness of signal features. Then, perform data preprocessing on the motor operation data information. After data preprocessing, use wavelet transform technology to extract the time-frequency features of the motor operation data information to generate the feature vector during motor operation. The feature vector during motor operation includes the magnetic flux density value in the air-gap and the vibration amplitude .

[0072] In this embodiment, first, through the deployment of the sensor group in step S1, an efficient monitoring network can be formed inside the permanent magnet generator. The sensor group includes various sensors for capturing vibration signals and magnetic field signals, and real-time operation data can be obtained during the operation of the motor, ensuring the comprehensiveness and accuracy of data collection. The layout of this sensor group further enhances the control of the motor operation state, and the real-time data information provides high-quality basic data support for subsequent data analysis. Secondly, by removing the noise and trend components in the data through wavelet denoising and moving average method in step S2, the stability and representativeness of the signal features can be ensured, and the reliability of the data is improved. The data preprocessing process eliminates the interference factors in the signal, making the characteristics of the air-gap eccentricity fault more prominent and clear. Based on the preprocessed data, wavelet transform is used to extract time-frequency features to generate feature vectors, which accurately describe the change of magnetic flux density and vibration amplitude in the air gap, providing multi-dimensional and refined feature information for subsequent eccentricity fault identification, and enhancing the fault identification ability of the system. Through these steps, this method further realizes the accurate monitoring of the motor operation state and multi-feature fusion, ensures the high accuracy of air-gap eccentricity fault identification, thereby improving the comprehensiveness and real-time performance of permanent magnet generator fault diagnosis, and providing effective early warning and maintenance guarantee for the stable operation of the motor.

[0073] Embodiment 3

[0074] Please refer to Figure 1 , specifically: The specific steps of S3 include:

[0075] S31. According to the feature vector during the operation of the motor, after feature extraction, analyze the change of the magnetic flux density in the air gap of the permanent magnet generator over time to calculate and obtain the change rate of magnetic field strength in different frequency bands , the change rate of magnetic field strength in different frequency bands is obtained through the following formula:

[0076] ;

[0077] In the formula, represents the magnetic flux density value in the air gap, represents the time interval, is the derivative of the magnetic flux density with respect to time, and this derivative value reflects its fluctuation, with the unit of Tesla per second (T / s); the magnetic field strength can be measured by a Hall sensor or a magnetic flux density sensor, and these sensors can capture the change of the magnetic field strength in the air gap over time.

[0078] The above-mentioned magnetic flux density value in the air gap reflects the strength of the internal magnetic field of the permanent magnet generator. In order to measure the magnetic flux density, it can be monitored and obtained through a Hall sensor or a magnetic flux density sensor;

[0079] S32. After feature extraction based on the eigenvector during the operation of the motor, analyze the synchronization between the vibration signal and the magnetic field signal of the permanent magnet generator to calculate and obtain the phase difference between the vibration signal and the magnetic field signal in different frequency bands , which is obtained specifically according to the following formula:

[0080] ;

[0081] In the formula, represents the vibration amplitude, represents the vibration amplitude and the dot product of the magnetic flux density value in the air gap , which is used to calculate the phase relationship between the two. The dot product is a scalar of the interaction between two signals, reflecting the amplitudes and relative directions of the two signals; represents the modulus of the vibration amplitude , represents the modulus of the magnetic flux density value in the air gap, is the arccosine function (also known as inverse cosine or arccosine value).

[0082] The above-mentioned vibration amplitude reflects the possible mechanical vibration or imbalance during the operation of the permanent magnet generator, which can be monitored and obtained through vibration sensors, accelerometers or piezoelectric sensors.

[0083] When using wavelet transform for time-frequency feature extraction of vibration signals and magnetic field signals, the signals will be decomposed into different frequency bands (i.e., different scales or frequency ranges). In each frequency band, there may be an independent phase relationship between the vibration signal and the magnetic field signal, so multiple sets of phase differences will be generated, and each set corresponds to a specific frequency band; the phase differences of multiple frequency bands extracted by wavelet transform can provide the relative phase information between the vibration signal and the magnetic field signal in different frequency ranges, providing key data support for the identification of air gap eccentricity and other faults. The existence of these multiple sets of phase differences enables the system to more comprehensively capture and analyze the complex characteristics of the motor in the fault state.

[0084] In this embodiment, first, through the feature extraction and magnetic flux density change rate calculation in step S31, the change of the magnetic field intensity in the air gap of the permanent magnet generator can be analyzed in detail. The change rate of the magnetic field intensity in different frequency bands reflects the characteristics of the air gap eccentricity fault at each frequency, which can help identify the specific location and spectral characteristics of the fault. This method captures the magnetic field change in real time by using Hall sensors or magnetic flux density sensors and calculates the time derivative of the magnetic flux density, which improves the sensitivity of the system to the uneven distribution of the magnetic field in the air gap and effectively enhances the ability to identify the early signs of the fault. Second, through the phase difference calculation in step S32, this method further analyzes the synchronism of the vibration signal and the magnetic field signal in different frequency bands; the phase difference between the vibration amplitude and the magnetic flux density provides the relative position information of the signal in the frequency domain, which can not only further help identify the type of air gap eccentricity (such as static or dynamic eccentricity), but also accurately capture the interaction between the vibration signal and the magnetic field signal. Using these phase characteristics, the system can more comprehensively reflect the unique patterns of the fault signal in each frequency band and effectively distinguish different types of eccentricity faults. This phase difference-based analysis method improves the reliability of fault diagnosis and provides an accurate positioning basis for subsequent fault troubleshooting and repair.

[0085] Embodiment 4

[0086] Please refer to Figure 1 , specifically: The specific steps of S3 also include:

[0087] S33. Based on the magnetic field intensity change rate obtained in S31 and the phase difference between the vibration signal and the magnetic field signal in different frequency bands , preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator, and after linear normalization processing, construct an eccentricity degree index Pczb. The eccentricity degree index Pczb is obtained through the following formula:

[0088] ;

[0089] In the formula, represents the number of frequency bands, , represents the magnetic field intensity change rate in the th frequency band, represents the average magnetic field intensity change rate, represents the phase difference between the vibration signal and the magnetic field signal in the th frequency band, represents the average value of the phase difference between the vibration signal and the magnetic field signal, and are both weight values, represents the first correction constant, where, and The specific values are set by the user according to the situation.

[0090] The specific steps of S3 also include:

[0091] S34. Preset a judgment threshold P, and by comparing and analyzing it with the eccentricity degree index Pczb, to judge whether there is a deviation in the air gap thickness of the current permanent magnet generator. The specific analysis content is as follows:

[0092] If the eccentricity degree index Pczb exceeds the judgment threshold P, it is judged that there is a deviation in the air gap thickness of the current permanent magnet generator (that is, there is a problem of air gap eccentricity fault), and at this time, a fault type troubleshooting instruction will be sent outwards;

[0093] If the eccentricity degree index Pczb does not exceed the judgment threshold P, it is judged that there is no deviation in the air gap thickness of the current permanent magnet generator for the time being, and at this time, a green warning notice will be sent outwards.

[0094] In this embodiment, first, in step S33, based on the magnetic field strength change rate and the phase difference between the vibration signal and the magnetic field signal in different frequency bands, the deviation of the air gap thickness of the permanent magnet generator can be accurately analyzed, and the eccentricity degree index Pczb is constructed through linear normalization processing. The calculation method of the eccentricity degree index adopts multi-frequency band signal analysis, and the mean values of the magnetic field strength change rate and the phase difference are integrated by weighting, ensuring the effective utilization of the information in each frequency band in fault identification, thereby improving the sensitivity of fault detection. The multi-frequency band analysis method of this method can more carefully reflect the complex characteristics of air gap eccentricity, making the eccentricity detection more accurate. Secondly, in step S34, by comparing the eccentricity degree index with a preset judgment threshold, it can intelligently judge whether there is a deviation in the air gap thickness of the current permanent magnet generator, realizing an automatic fault judgment and feedback mechanism. When the eccentricity degree index exceeds the threshold, the system automatically issues a fault type troubleshooting instruction, which helps to conduct detailed fault analysis in a timely manner; when it does not exceed the threshold, the system only issues a green warning notice, indicating that the equipment is in normal operation. This dual-mode notification mechanism of intelligent warning and fault troubleshooting not only improves the real-time performance of fault identification, but also optimizes the response efficiency of the system, further avoiding false alarms or missed alarms, and providing a reliable guarantee for the stable operation of the permanent magnet generator.

[0095] Embodiment 5

[0096] Please refer to Figure 1 , specifically: The specific steps of S4 include:

[0097] S41, after receiving the fault type troubleshooting instruction issued in S34, using a thermal imaging device to capture an infrared thermal image of the air gap area in the permanent magnet generator, and using Gaussian filtering technology to perform image denoising on the infrared thermal image to reduce background and noise interference, wherein the thermal imaging device refers to an infrared thermal imager;

[0098] S42, acquiring infrared thermal data according to the infrared thermal image after image denoising, wherein the infrared thermal data includes the temperature value at each pixel position in the infrared thermal image ;

[0099] S43, by calculating the temperature value at each pixel position in the infrared thermal image Compare with the preset threshold value, if the temperature value at each pixel position in the infrared thermal image When the value exceeds the preset threshold, the corresponding pixel position is marked as an outlier, and the outlier is counted to obtain the number of outliers YS.

[0100] The preset threshold value can be determined by the statistical characteristics of the temperature distribution of the image (such as mean value, standard deviation), specifically: preset threshold value = temperature mean value + k* temperature standard deviation; k is a constant, usually with a value of 1-3, corresponding to different confidence levels, and the specific value is set by the user (according to the actual situation);

[0101] The specific steps of S4 also include:

[0102] S43, analyzing the risk of foreign matter in the air gap area of the permanent magnet generator according to the infrared thermal data, so as to construct the temperature gradient at each pixel position , which can be obtained by:

[0103] ;

[0104] In the formula, Indicates pixel position The temperature gradient at a location can also be understood as the local temperature change rate at that location. The larger the gradient value, the more drastic the temperature change around that location, which usually occurs in temperature mutation areas (such as hot spot edges). It is expressed as the temperature gradient along the horizontal axis, that is, the rate of change of temperature in the x direction. This is a partial derivative that measures the pixel position The temperature variation relative to its adjacent points in the x direction; It is expressed as the temperature gradient along the vertical axis, that is, the rate of change of temperature in the y direction. This is a partial derivative that measures the pixel position The temperature change in the y direction relative to its adjacent points; x and y represent the horizontal and vertical coordinates in the infrared thermal image, respectively.

[0105] In this embodiment, first, in step S41, an infrared thermal image of the air gap region inside the permanent magnet generator is captured by a thermal imaging device, and Gaussian filtering technology is used for image denoising, effectively reducing background and noise interference, ensuring the clarity and accuracy of the infrared thermal image. This denoising process makes the temperature data in the image more representative, ensuring the accuracy of temperature analysis and laying a high-quality data foundation for subsequent fault analysis. Secondly, in steps S42 and S43, by comparing the temperature values of each pixel position in the infrared thermal image with a preset threshold, the pixels exceeding the threshold are marked as abnormal points and the number of abnormal points is counted. This threshold setting method based on the temperature mean and standard deviation improves the recognition accuracy and can adapt to the temperature distribution characteristics under different working environments, making the fault recognition more robust. In addition, the temperature gradient calculated based on the infrared thermal data in S43 can reflect the local temperature change situation at each position in the air gap region. The calculation of the temperature gradient provides a detailed distribution of the temperature change in the air gap region. The larger the gradient, the more likely it indicates the presence of foreign objects in the air gap, thus improving the positioning accuracy of hot spots and foreign objects in the air gap. In summary, through this method, local hot spots caused by foreign objects in the air gap region of the permanent magnet generator can be further effectively identified, making the fault recognition more accurate and comprehensive, providing a strong guarantee for the detection of foreign objects in the air gap and the safety of the operating state of the permanent magnet generator. While improving the accuracy and stability of air gap foreign object detection, this method reduces the occurrence of false alarms and further optimizes the operation management of the permanent magnet generator.

[0106] Embodiment 6

[0107] Please refer to Figure 1 , specifically: The specific steps of S5 include:

[0108] S51. Compare and analyze the temperature gradient at each position with the safety threshold Q to determine the fault type of the current permanent magnet generator. The specific judgment content is as follows:

[0109] S511. If the temperature gradient at the corresponding position exceeds the safety threshold Q, mark the current position as a local hot spot at this time;

[0110] S512. If the temperature gradient at the corresponding position does not exceed the safety threshold Q, do not mark the current position as a local hot spot for the time being;

[0111] S52. Count the local hot spots obtained in S511 to obtain the number of local hot spots JS. When the number of local hot spots JS exceeds 80% of the number of abnormal points YS, it is determined at this time that the current permanent magnet generator has a foreign object fault in the air gap.

[0112] In this embodiment, first, in step S51, by comparing the temperature gradients at each position with a preset safety threshold, the system can effectively determine local hotspots with abnormal temperatures; the positions where the temperature gradients exceed the safety threshold are automatically marked as local hotspots, which helps to identify the temperature concentration phenomenon in the air gap area in real time, improving the response speed to air gap eccentricity and foreign object faults. Secondly, in step S52, the system counts the marked local hotspots, obtains the number of local hotspots, and compares and analyzes it with the total number of abnormal points. When the number of local hotspots exceeds 80% of the number of abnormal points, the system determines that there is an air gap foreign object fault. This analysis method based on statistical comparison can accurately identify the local temperature abnormality caused by foreign objects in the air gap, further avoiding misjudgment and missed judgment. This method realizes high-sensitivity detection of air gap foreign object faults through the comparison of the number of local hotspots and the number of abnormal points, providing data support for the accurate judgment of fault types.

[0113] Embodiment 7

[0114] Please refer to Figure 2 , specifically: A permanent magnet generator air gap eccentricity fault identification system includes an acquisition module, a processing module, an eccentricity analysis module, a type troubleshooting module, and a type feedback module;

[0115] The acquisition module is used to pre-deploy several groups of sensors inside the permanent magnet generator and based on the several groups of sensors, monitor the vibration signal and magnetic field signal of the permanent magnet generator during operation in real time to construct motor operation data information;

[0116] The processing module is used to perform data preprocessing on the motor operation data information and use wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal to generate a feature vector during motor operation;

[0117] The eccentricity analysis module is used to preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator based on the feature vector formed after data preprocessing to construct an eccentricity degree index Pczb, and based on the value of the eccentricity degree index Pczb, determine whether to send out a fault type troubleshooting instruction externally;

[0118] The type troubleshooting module is used to, when receiving the fault type troubleshooting instruction, use thermal imaging technology to monitor the two-dimensional temperature distribution in the air gap area of the permanent magnet generator to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator and construct the temperature gradient at each position ;

[0119] The type feedback module is used to preset a safety threshold Q, by the temperature gradient at each position Compare and analyze with the safety threshold Q to determine the fault type of the current permanent magnet generator.

[0120] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying the air-gap eccentricity fault of a permanent magnet generator, characterized in that: Including the following steps, S1. Deploy several groups of sensors inside the permanent magnet generator in advance, and based on the several groups of sensors, real-time monitor the vibration signal and magnetic field signal of the permanent magnet generator during operation to construct motor operation data information; S2. Through data preprocessing of the motor operation data information and using wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal, generate a feature vector during motor operation; S3. Based on the feature vector formed after data preprocessing in S2, preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator to construct an eccentricity degree index Pczb. Based on the value of the eccentricity degree index Pczb, judge whether to send out a fault type troubleshooting instruction; S4. After receiving the fault type troubleshooting instruction, use thermal imaging technology to monitor the two-dimensional temperature distribution in the air gap area of the permanent magnet generator to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator, and construct the temperature gradient at each position. ; S5. Preset a safety threshold Q, and by comparing and analyzing the temperature gradient at each position with the safety threshold Q, determine the fault type of the current permanent magnet generator.

2. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 1, characterized in that: The specific steps of S1 include: S11. Deploy several groups of sensors inside the permanent magnet generator in advance to form a sensor group. The sensor group includes vibration sensors and Hall sensors; S12. According to the sensor group formed in S11, real-time monitor and obtain the vibration signal and magnetic field signal of the permanent magnet generator during operation to construct motor operation data information.

3. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 2, characterized in that: The specific steps of S2 include: S21. Remove the noise in the motor operation data information through wavelet denoising technology, and remove the trend component in the motor operation data information through the moving average method to perform data preprocessing operations on the motor operation data information. After data preprocessing, use the wavelet transform technology to extract the time-frequency characteristics of the motor operation data information to generate a feature vector during motor operation. The feature vector during motor operation includes the magnetic flux density value in the air gap and the vibration amplitude .

4. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 3, characterized in that: The specific steps of S3 include: S31. After feature extraction based on the eigenvector during the operation of the motor, analyze the change of the magnetic flux density in the air gap of the permanent magnet generator over time to calculate and obtain the change rate of the magnetic field strength in different frequency bands. The change rate of the magnetic field strength in different frequency bands is obtained through the following formula: ； wherein, represents the magnetic flux density value in the air gap, represents the time interval, is the derivative of the magnetic flux density with respect to time; S32. After feature extraction based on the eigenvector during the operation of the motor, analyze the synchronism between the vibration signal and the magnetic field signal of the permanent magnet generator to calculate and obtain the phase difference between the vibration signal and the magnetic field signal in different frequency bands , which is obtained specifically according to the following formula: ； In the formula, represents the vibration amplitude, represents the vibration amplitude and the dot product of the magnetic flux density value in the air gap is used to calculate the phase relationship between the two. The dot product is a scalar of the interaction between two signals, reflecting the amplitudes and relative directions of the two signals; represents the vibration amplitude modulus of, represents the magnetic flux density value in the gap modulus of, is the inverse cosine function.

5. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 4, characterized in that: The specific steps of S3 also include: S33. Based on the magnetic field intensity change rate obtained in S31 and S32 and the phase difference between the vibration signal and the magnetic field signal in different frequency bands , preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator, and after linear normalization processing, construct an eccentricity index Pczb. The eccentricity index Pczb is obtained through the following formula: ； Wherein, represents the number of frequency bands, , represents the rate of change of magnetic field strength on the th frequency band, represents the average rate of change of magnetic field strength, represents the phase difference between the vibration signal and the magnetic field signal on the th frequency band, represents the average value of the phase difference between the vibration signal and the magnetic field signal, and are both weight values, represents the first correction constant, where and The specific values are set by the user according to the situation.

6. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 5, characterized in that: The specific steps of S3 also include: S34. Preset a judgment threshold P, and by comparing and analyzing it with the eccentricity degree index Pczb, judge whether there is a deviation in the air gap thickness of the current permanent magnet generator. The specific analysis content is as follows: If the eccentricity degree index Pczb exceeds the judgment threshold P, it is judged that there is a deviation in the air gap thickness of the current permanent magnet generator, and at this time, a fault type troubleshooting instruction will be sent out; If the eccentricity degree index Pczb does not exceed the judgment threshold P, it is judged that there is no deviation in the air gap thickness of the current permanent magnet generator for the time being, and at this time, a green warning notice will be sent out.

7. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 6, characterized in that: The specific steps of S4 include: S41. After receiving the fault type troubleshooting instruction issued in S34, use a thermal imaging device to capture the infrared thermal image of the air gap area inside the permanent magnet generator, and use Gaussian filtering technology to denoise the infrared thermal image. Among them, the thermal imaging device refers to an infrared thermal imager; S42. Obtain infrared thermal data based on the infrared thermal image after image denoising, where the infrared thermal data includes temperature values at each pixel position in the infrared thermal image. ; S43. By comparing the temperature values at each pixel position in the infrared thermal image with a preset threshold value, if the temperature values at each pixel position in the infrared thermal image exceed the preset threshold value, then mark the corresponding pixel positions as abnormal points, and count the abnormal points to obtain the number of abnormal points YS.

8. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 7, characterized in that: The specific steps of S4 also include: S43. Analyze the risk of air-gap foreign objects in the air-gap region of the permanent magnet generator based on the infrared thermal data to construct the temperature gradient at each pixel position , which is obtained specifically in the following manner: ； In the formula, represents the temperature gradient in the horizontal axis direction; represents the temperature gradient in the vertical axis direction; x and y respectively represent the horizontal and vertical coordinates in the infrared thermal image.

9. A method for identifying the air-gap eccentricity fault of a permanent magnet generator according to claim 7, characterized in that: The specific steps of S5 include: S51. Compare the temperature gradient at each position with the safety threshold Q for comparison and analysis to determine the fault type of the current permanent magnet generator. The specific judgment content is as follows: S511. If the temperature gradient at the corresponding position exceeds the safety threshold Q, the current position is marked as a local hot spot at this time; S512. If the temperature gradient at the corresponding position does not exceed the safety threshold Q, the current position is not marked as a local hot spot at this time; S52. Statistically analyze the local hot spots obtained in S511 to obtain the number of local hot spots JS. When the number of local hot spots JS exceeds 80% of the number of abnormal points YS, it is judged that there is an air gap foreign object fault in the current permanent magnet generator.

10. A permanent magnet generator air-gap eccentricity fault identification system for implementing the permanent magnet generator air-gap eccentricity fault identification method according to any one of claims 1 to 9 above, characterized in that: Including an acquisition module, a processing module, an eccentricity analysis module, a type troubleshooting module, and a type feedback module; The acquisition module is used to deploy several groups of sensors inside the permanent magnet generator in advance, and based on the several groups of sensors, real-time monitor the vibration signal and magnetic field signal of the permanent magnet generator during operation to construct motor operation data information; The processing module is used to perform data preprocessing on the motor operation data information and use wavelet transform to obtain the time-frequency characteristics of the vibration signal and magnetic field signal to generate a feature vector during motor operation; The eccentric analysis module is used to preliminarily analyze whether there is a deviation in the air gap thickness of the permanent magnet generator based on the feature vector formed after data preprocessing, so as to construct an eccentric degree index Pczb. Based on the value of the eccentric degree index Pczb, it is judged whether it is necessary to send out a fault type troubleshooting instruction outward; The type troubleshooting module is used to, after receiving a fault type troubleshooting instruction, utilize thermal imaging technology to monitor the two-dimensional temperature distribution in the air gap area of the permanent magnet generator, so as to obtain infrared thermal data, and based on the infrared thermal data, analyze the risk of air gap foreign objects in the air gap area of the permanent magnet generator, and construct the temperature gradient at each position ; The type feedback module is used to preset a safety threshold Q. By comparing the temperature gradient at each position with the safety threshold Q for comparison and analysis, the fault type of the current permanent magnet generator is determined.

Images