A test method for diagnosing eccentricity fault of asynchronous motor

Abstract

The present application discloses a test method for diagnosing an asynchronous motor eccentric fault, and relates to the technical field of motor fault diagnosis. The method includes: obtaining historical fault data of an asynchronous motor eccentric fault, extracting eccentric fault characteristics; diagnosing whether an asynchronous motor has an eccentric fault, and outputting primary fault data if an eccentric fault occurs; obtaining current data of the asynchronous motor and the corresponding detection environment to form current spectrum data; testing whether an asynchronous motor has an eccentric fault and outputting a test result; if the test result is that an asynchronous motor has an eccentric fault, obtaining a monitoring image of the asynchronous motor during operation, and extracting vibration data of the asynchronous motor in combination with an attention mechanism; obtaining vibration modes of different components of the asynchronous motor, and confirming the eccentric position and severity of the eccentricity in combination with the vibration data to obtain a fault result. The present application improves the accuracy of the test for diagnosing an asynchronous motor eccentric fault.

Description

Technical Field

[0001] The present application relates to the technical field of motor fault diagnosis, and in particular to a test method for diagnosing eccentricity faults of asynchronous motors. Background Art

[0002] Motors appear in all aspects of daily life. They are found in cars and various large equipment. The use of motors provides convenience for human life. In motors, eccentricity refers to the misalignment between the rotor (or rotating) center and the stator center. Eccentricity can have a significant impact on the performance of the motor. If it is not resolved, it will lead to increased vibration, reduced efficiency and even motor failure. Therefore, it is particularly important to diagnose motor eccentricity faults and obtain accurate eccentricity fault results through inspection.

[0003] In the patent with publication number CN113093005A, entitled "A method for diagnosing static eccentricity faults of permanent magnet synchronous motors", the diagnosis of motor eccentricity faults relies on electromagnetic parameter analysis to obtain diagnostic results. This not only requires the arrangement of detection coils to increase external influences during the operation of the motor, but also the electromagnetic field is easily affected by external interference, thus resulting in inaccurate motor eccentricity fault diagnosis results and adverse effects. Summary of the invention

[0004] The object of the present invention is to provide a method for diagnosing an asynchronous motor eccentricity fault, so as to solve the problems raised in the above-mentioned background technology.

[0005] The present application provides a method for diagnosing an asynchronous motor eccentricity fault, which adopts the following technical solution:

[0006] Acquire historical fault data of asynchronous motor eccentricity fault, and extract eccentricity fault features from the historical fault data;

[0007] Diagnose whether the asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics, and output primary fault data if an eccentricity fault occurs;

[0008] Obtain the current data of the asynchronous motor and the corresponding detection environment, perform quality screening on the current data according to the detection environment, and form current spectrum data;

[0009] According to the current spectrum data, the frequency and amplitude in the current spectrum data are extracted, and according to the frequency and amplitude, whether the asynchronous motor has an eccentricity fault is detected and the detection result is output;

[0010] If the inspection result shows that the asynchronous motor has an eccentric fault, the monitoring image of the asynchronous motor during operation is obtained, and the vibration data of the asynchronous motor is extracted by combining the attention mechanism;

[0011] The vibration modes of different parts of the asynchronous motor are obtained, and the eccentricity position and severity are confirmed by combining the vibration data to obtain the fault results.

[0012] Preferably, the step of diagnosing whether the asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics and outputting primary fault data if an eccentricity fault occurs is specifically:

[0013] Statistically calculate the fault frequency AL of different eccentricity fault characteristics in the historical fault data of asynchronous eccentricity faults;

[0014] Obtain historical fault data of all fault types of asynchronous motors, and count the number of fault types AS covered by the eccentricity fault characteristics;

[0015] The fault eccentricity AP is calculated according to the fault eccentricity formula AP = a1 × AL-a2 × AS, where a1 and a2 are proportional factors and are greater than 0;

[0016] Find the fault eccentricity corresponding to the eccentricity fault feature, superimpose the fault eccentricity of all extracted eccentricity fault features, and obtain the eccentricity fault probability;

[0017] An eccentricity fault probability threshold is set to determine whether the eccentricity fault probability reaches the eccentricity fault probability threshold. If so, the asynchronous motor is diagnosed to have an eccentricity fault and primary fault data is output.

[0018] Preferably, the step of acquiring the current data of the asynchronous motor and the corresponding detection environment, and performing quality screening on the current data according to the detection environment to form the current spectrum data is specifically as follows:

[0019] Acquire the current data of the asynchronous motor and the detection environment during the current data detection to obtain real-time detection data;

[0020] Obtain standard use environment data of the current detection device, compare and obtain the difference between the real-time detection data and the standard use environment data and record it as the environment difference;

[0021] Set an environmental difference threshold, remove current data whose environmental difference reaches the environmental difference threshold, and record the removal time point of the removed current data;

[0022] Set a time period, intercept the current data in the time period before the elimination time point, form a correlation function according to the change of the current data, and calculate the estimated current data at the elimination time point according to the correlation function;

[0023] The predicted current data replaces the eliminated current data to form new current data, and the parameters of the current data are extracted to form current spectrum data.

[0024] Preferably, the step of extracting the frequency and amplitude in the current spectrum data according to the current spectrum data, and inspecting whether the asynchronous motor has an eccentricity fault according to the frequency and amplitude and outputting the inspection result is specifically:

[0025] Get the actual running speed n of the motor, the synchronous speed n1 and 0≤n≤n1, according to the slip formula The slip rate s is calculated;

[0026] Obtain relevant parameters of the motor, extract the main parameters affecting the current spectrum, and obtain the empirical formula for air gap eccentricity Where f1 is the power supply frequency, s is the slip rate, p is the number of magnetic pole pairs, and k is an arbitrary integer;

[0027] The electromagnetic torque is calculated according to the torque formula Where m1 is the number of stator phases, p is the number of magnetic pole pairs, U1 is the stator phase voltage, f1 is the power supply frequency, r1 and x1 are the resistance and leakage reactance of each phase winding of the stator, and r′2 and x′2 are the rotor resistance and leakage reactance converted to the stator side;

[0028] The actual operating speed of the asynchronous motor is obtained according to the correlation function of the motor speed and the number of magnetic pole pairs n=60f / p;

[0029] The standard fault range corresponding to the frequency and amplitude is obtained according to the actual operating speed, and the asynchronous motor is tested for eccentricity fault and the test results are output.

[0030] Preferably, the step of obtaining the standard fault range corresponding to the frequency and amplitude according to the actual operating speed, inspecting whether the asynchronous motor has an eccentricity fault and outputting the inspection result is specifically:

[0031] Generate a mechanical characteristic curve according to the electromagnetic torque, obtain the value range of the slip rate s, and obtain the maximum and minimum values ​​of the slip rate s;

[0032] Get the value of p, and discuss the different values ​​of p by category. When the value of p is fixed, the maximum value of slip rate s, the minimum value of slip rate s, and the value of k are used as variables for cross analysis to form a fault table.

[0033] According to the frequency in the fault table, the ratio of the amplitude of the side frequency component to the fundamental frequency component corresponding to the frequency is calculated, and the ratio range of the amplitude of the side frequency component to the fundamental frequency component is obtained;

[0034] Extract the frequency and amplitude from the current spectrum data, determine whether the frequency meets the corresponding frequency in the fault table, and whether the ratio of the corresponding side frequency component to the fundamental frequency component amplitude is within the ratio range;

[0035] If the frequency meets the corresponding frequency in the fault table, and the ratio of the corresponding side frequency component to the fundamental frequency component amplitude is within the ratio range, it is determined that the asynchronous motor has an eccentricity fault and the test result is output.

[0036] Preferably, if the inspection result is that the asynchronous motor has an eccentricity fault, the monitoring image of the asynchronous motor when running is obtained, and the vibration data of the asynchronous motor is extracted by combining the attention mechanism, specifically, the step of:

[0037] Extracting fault vibration data of eccentricity fault from historical fault data of asynchronous motor eccentricity fault, and marking the eccentricity position of fault vibration data;

[0038] Obtain the normal vibration data of the asynchronous motor when it is working normally, and put the fault vibration data and normal vibration data into the attention mechanism model training;

[0039] Obtain monitoring images of the asynchronous motor during operation, and obtain real-time vibration data of various parts of the asynchronous motor based on the monitoring image analysis;

[0040] The real-time vibration data is input into the trained attention mechanism model, and the data associated with the asynchronous motor eccentricity fault is filtered out through the attention mechanism and the vibration data is output.

[0041] Preferably, the step of acquiring a monitoring image of the asynchronous motor when it is running and obtaining real-time vibration data of various parts of the asynchronous motor according to the monitoring image analysis is specifically:

[0042] Continuously collect monitoring images of the asynchronous motor when it is running, and extract characteristic information of the asynchronous motor in the image after preprocessing the monitoring images;

[0043] The changes of characteristic information in the monitoring image sequence are obtained, and the vibration displacement and vibration velocity are obtained according to the changes, so as to output the real-time vibration data of each part of the asynchronous motor.

[0044] Preferably, the step of obtaining the vibration modes of different components of the asynchronous motor, confirming the eccentricity position and the severity of the eccentricity in combination with the vibration data, and obtaining the fault result is specifically:

[0045] Obtain the vibration position corresponding to the most relevant vibration data in the attention mechanism as the eccentric position of the asynchronous motor;

[0046] Obtain a connection drawing of the asynchronous motor, and calculate the distance between the asynchronous eccentric position and each component of the asynchronous motor according to the connection drawing and the asynchronous eccentric position and record it as the eccentric distance;

[0047] Obtain the impact distance of the eccentricity fault on other components, find the components of the asynchronous motor whose eccentricity distance is within the impact distance and record them as the impact components;

[0048] The severity of eccentricity is obtained based on the resonance probability and functional importance of the affected components;

[0049] The fault result is output based on the eccentricity position and eccentricity severity.

[0050] Preferably, the step of obtaining the severity of eccentricity according to the resonance probability and functional importance of the influencing component is specifically:

[0051] Obtain the component vibration mode of the affected component, obtain the eccentric vibration mode of the asynchronous motor eccentricity fault, compare the component vibration mode and the eccentric vibration mode, and obtain the modal similarity;

[0052] Obtain the standard similarity of modal resonance, select the affected components whose modal similarity reaches the standard similarity and record them as faulty components;

[0053] Obtain the importance of the function of the faulty component BG, and obtain the influence of the faulty component on the asynchronous motor BY;

[0054] The modal similarity BX corresponding to the faulty component is obtained, and the eccentricity severity BP of the faulty component is calculated according to the eccentricity severity correlation function BP=b1×BG+b2×BY+b3×BX, where b1, b2, and b3 are proportional factors and are greater than 0.

[0055] In summary, the present application includes at least one of the following beneficial technical effects:

[0056] 1. After the preliminary diagnosis of the eccentricity fault of the asynchronous motor, the current frequency and the range of the ratio of the amplitude to the fundamental frequency amplitude during the eccentricity fault are determined. After the quality screening of the current data, the current spectrum formed by the current data is compared to see whether it is within the range of the eccentricity fault. The eccentricity fault is further confirmed and a more accurate eccentricity fault result is obtained, which improves the accuracy of the test of the eccentricity fault diagnosis of the asynchronous motor.

[0057] 2. Through the training of the attention mechanism model, the monitoring image pays more attention to the fault situation and the detection of effective data. After preprocessing the detected image, the changes in the feature information in the sequence are observed to obtain the vibration data of the asynchronous motor. Through the attention mechanism detection, the processing of invalid data is reduced and the speed of data processing is improved. Vibration data is obtained through image detection, which reduces the interference to the asynchronous motor and improves the convenience of the inspection of asynchronous motor eccentricity fault diagnosis.

[0058] 3. The eccentricity position is confirmed through vibration data, and the affected components are screened according to the distance between the eccentricity position and each component. Then, the consequences of the eccentricity fault are obtained and the severity of the eccentricity fault is evaluated by combining the similarity between the component vibration mode and the fault vibration mode, the functional importance of the component, and the impact of the component. More accurate fault data is obtained, which improves the accuracy and convenience of the inspection of asynchronous motor eccentricity fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the specific steps of an embodiment of a test method for diagnosing an eccentricity fault of an asynchronous motor according to the present invention. DETAILED DESCRIPTION

[0060] Below is a combination of the embodiments and Figure 1 The present invention is further described in detail, but the embodiments of the present invention are not limited thereto.

[0061] The invention discloses a test method for diagnosing an asynchronous motor eccentricity fault, which specifically comprises the following steps:

[0062] Step S1, obtaining historical fault data of asynchronous motor eccentricity fault, and extracting eccentricity fault characteristics from the historical fault data.

[0063] The fault characteristics of eccentricity include vibration, noise, magnetic field changes, etc.

[0064] Step S2, diagnose whether the asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics, and output primary fault data if an eccentricity fault occurs.

[0065] Step S3, obtaining the current data of the asynchronous motor and the corresponding detection environment, and performing quality screening on the current data according to the detection environment to form current spectrum data.

[0066] Step S4, extracting the frequency and amplitude in the current spectrum data according to the current spectrum data, and checking whether the asynchronous motor has an eccentricity fault according to the frequency and amplitude, and outputting the test result.

[0067] Step S5: If the inspection result shows that the asynchronous motor has an eccentricity fault, a monitoring image of the asynchronous motor during operation is obtained, and vibration data of the asynchronous motor is extracted by combining the attention mechanism.

[0068] Step S6, obtaining the vibration modes of different components of the asynchronous motor, confirming the eccentricity position and the severity of the eccentricity in combination with the vibration data, and obtaining the fault result.

[0069] In actual use, there are many types of asynchronous motor faults, and many faults will cause some parameters of the asynchronous motor to change. Therefore, the eccentricity fault of the asynchronous motor is not easy to diagnose accurately. After diagnosing the asynchronous motor fault, the fault diagnosis of the asynchronous motor is tested again, which is conducive to obtaining more accurate fault results. At the same time, testing the eccentricity position and severity of the asynchronous motor is conducive to formulating a more appropriate and accurate maintenance plan. For example, when it is initially judged as an eccentricity fault based on the fault characteristics, it is verified whether the eccentricity fault has indeed occurred based on the current spectrum signal. If an eccentricity fault has indeed occurred, the eccentricity position and severity of the eccentricity are determined based on the vibration data, and the maintenance personnel are notified in time to perform maintenance.

[0070] The steps of diagnosing whether an asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics and outputting primary fault data if an eccentricity fault occurs are as follows:

[0071] Step S21 , counting the fault frequencies AL of different eccentricity fault characteristics in the historical fault data of asynchronous eccentricity faults.

[0072] Step S22, obtaining historical fault data of all fault types of the asynchronous motor, and counting the number AS of fault types covered by the eccentricity fault feature.

[0073] Step S23, calculating the fault eccentricity AP according to the fault eccentricity formula AP=a1×AL-a2×AS, wherein a1 and a2 are proportional factors and are greater than 0.

[0074] Step S24, searching for the fault eccentricity corresponding to the eccentricity fault feature, superimposing the fault eccentricities of all extracted eccentricity fault features, and obtaining the eccentricity fault probability.

[0075] Step S25, setting an eccentricity fault probability threshold, determining whether the eccentricity fault probability reaches the eccentricity fault probability threshold, and if so, diagnosing that the asynchronous motor has an eccentricity fault and outputting primary fault data.

[0076] In actual use, there are many types of asynchronous motor faults. The changes of many parameters of asynchronous motors will be affected by the faults. When an asynchronous motor fails, it is necessary to determine whether the fault of the asynchronous motor is an eccentric fault. The fault frequency of the eccentric fault feature is counted. When an asynchronous eccentric fault occurs, the higher the frequency of the fault feature, the more likely it is that the feature is the main feature of the eccentric fault. The more fault types the fault covers, the more difficult it is for the feature to prove that the fault is an eccentric fault. For example, the noise of the eccentric fault feature becomes louder. When the asynchronous motor has broken rotor bars, loose cores, improper bearing installation and other fault types, the noise will become louder. Therefore, when the noise becomes louder, it is difficult to determine whether it is an eccentric fault. The air gap change occurs more frequently in eccentric faults, accounting for 100%, so the air gap change can better reflect that the fault is an eccentric fault. According to the frequency of occurrence of the fault feature and the overlap rate with other fault types, the probability of the fault type being an eccentric fault can be judged, thereby achieving the result of preliminary diagnosis of the fault.

[0077] The steps of obtaining the current data of the asynchronous motor and the corresponding detection environment, and performing quality screening on the current data according to the detection environment to form the current spectrum data are specifically as follows:

[0078] Step S31, acquiring the current data of the asynchronous motor and the detection environment during the current data detection, to obtain real-time detection data.

[0079] Step S32, obtaining standard use environment data of the current detection device, comparing and obtaining the difference between the real-time detection data and the standard use environment data and recording it as the environment difference.

[0080] Step S33, setting an environmental difference threshold, eliminating current data whose environmental difference reaches the environmental difference threshold, and recording the elimination time point of the eliminated current data.

[0081] Step S34, setting a time period, intercepting the current data in the time period before the elimination time point, forming a correlation function according to the change of the current data, and calculating the estimated current data at the elimination time point according to the correlation function.

[0082] Step S35, replacing the eliminated current data with the predicted current data to form new current data, and extracting parameters of the current data to form current spectrum data.

[0083] In actual use, the detection of current will also be affected by the external environment. For some current data affected by the external environment, if used as reference data for eccentric fault detection, it will affect the correctness of the eccentric fault detection results. Therefore, it is necessary to eliminate and replace the current data that does not meet the quality standards to improve the accuracy of eccentric fault detection. When performing current detection, if there is a strong magnetic field in the environment, it will affect the electronic components in the detection equipment, resulting in measurement errors. Changes in ambient humidity will affect the insulator resistance in the measuring tool, causing changes in capacitance characteristics, which in turn affects the accuracy of the measurement results. Temperature changes will cause changes in the resistance value inside the device, thereby affecting the measurement accuracy of the current. Including noise, interference, equipment pollution and other situations will affect the accuracy of current data, and current data that does not meet the quality standards needs to be eliminated and replaced. For example, at a certain point in time, the current data does not meet the standards due to noise. A correlation function is formed based on the current data of the previous time period, and the frequency of the current signal is steadily rising. The theoretical current data of this point is calculated according to the correlation function for replacement.

[0084] The steps of extracting the frequency and amplitude in the current spectrum data according to the current spectrum data, and inspecting whether the asynchronous motor has an eccentricity fault according to the frequency and amplitude and outputting the inspection result are specifically as follows:

[0085] Step S41, obtain the actual running speed n of the motor, the synchronous speed n1 and 0≤n≤n1, according to the slip formula The slip rate s is calculated.

[0086] The speed range of the asynchronous motor is 0≤n≤n1, and the corresponding slip rate is 0≤s≤1

[0087] Step S42, obtaining relevant parameters of the motor, extracting the main parameters affecting the current spectrum, and obtaining an empirical formula for air gap eccentricity Where f1 is the power supply frequency, s is the slip rate, p is the number of magnetic poles, and k is an arbitrary integer.

[0088] Step S43, calculating the electromagnetic torque according to the torque formula Where m1 is the number of stator phases, p is the number of pole pairs, U1 is the stator phase voltage, f1 is the power supply frequency, r1 and x1 are the resistance and leakage reactance of each phase winding of the stator, and r′2 and x′2 are the rotor resistance and leakage reactance converted to the stator side.

[0089] The output power of an asynchronous motor is closely related to the rotor speed.

[0090] Step S44, obtaining the actual operating speed of the asynchronous motor according to the correlation function n=60f / p between the motor speed and the number of magnetic pole pairs.

[0091] Step S45, obtaining the standard fault range corresponding to the frequency and amplitude according to the actual operating speed, checking whether the asynchronous motor has an eccentricity fault and outputting the test result.

[0092] In actual use, according to the movement of the motor when it is eccentric, the eccentricity fault can be divided into three types of eccentricity: static eccentricity, dynamic eccentricity and mixed eccentricity. Under normal circumstances, the axis of the motor stator, the axis of the rotor and the axis of the rotor rotation should coincide, but due to processing technology, installation and debugging and other reasons, the motor stator and rotor axis system do not coincide, which in turn causes eccentricity faults. Static eccentricity refers to the offset of the rotor axis, the stator axis remains in place, and the rotor rotates along its own geometric axis. At this time, the position of the minimum air gap does not change. Dynamic eccentricity refers to the position of the rotor axis remains unchanged, the rotation axis of the stator is offset, and the position of the minimum air gap is constantly changing. Mixed eccentricity includes static eccentricity and dynamic eccentricity. After the eccentricity fault occurs, the uneven air gap between the stator and the rotor causes the asymmetric magnetic circuit on the motor axis, and there are differences in the degree of saturation. Specifically, in the larger air gap of the motor, the air gap magnetic resistance is larger, the magnetic flux density on the magnetic circuit passing through this place is smaller, and it is not easy to saturate; on the contrary, in the smaller air gap, it is easier to saturate.

[0093] Based on the theoretical formula:

[0094]

[0095] f1 is the power supply frequency, s is the slip rate, R is the number of rotor bars of the squirrel cage asynchronous motor, p is the number of magnetic pole pairs, k1=0 for static eccentricity, k1=1 for dynamic eccentricity, k2=1, 3, 5…;

[0096] According to the group standard of the Institute of Electrical Engineering, the characteristic is most obvious when k2=1. The theoretical formula is used as the characteristic component for detecting air gap eccentricity fault. Since too many motor structural parameters are involved, diagnosis is extremely difficult when the parameters are difficult to obtain. After years of practical application, an empirical formula for air gap eccentricity is summarized.

[0097] The steps of obtaining the standard fault range corresponding to the frequency and amplitude according to the actual operating speed, checking whether the asynchronous motor has an eccentricity fault and outputting the test results are as follows:

[0098] Step S451, generating a mechanical characteristic curve according to the electromagnetic torque, obtaining a value range of the slip rate s, and obtaining a maximum value and a minimum value of the slip rate s.

[0099] When the motor speed decreases and the slip rate increases, the motor torque will first increase and then decrease. When the motor runs at rated power, it can maximize the motor operating efficiency under the safest and most stable conditions. At this time, the corresponding slip rate is s n When the motor operates at the critical slip rate S mUnder the working condition of , the motor can provide the maximum torque, but the motor cannot run for a long time under this working condition; In summary, when the motor is operating normally, the maximum slip rate that can be achieved is the critical slip rate S m ,After statistical analysis, the critical slip rate is between 0.05-0.06, so the maximum slip rate can be taken as 0.06 when calculating.

[0100] Step S452, obtaining the value of p, and classifying and discussing different values ​​of p. When the value of p is fixed, the maximum value of slip rate s, the minimum value of slip rate s, and the value of k are used as variables for cross analysis to form a fault table.

[0101] Step S453, according to the frequency in the fault table, calculate the ratio of the amplitude of the side frequency component to the amplitude of the fundamental frequency component corresponding to the frequency, and obtain the ratio range of the amplitude of the side frequency component to the amplitude of the fundamental frequency component.

[0102] Step S454, extracting the frequency and amplitude in the current spectrum data, determining whether the frequency satisfies the corresponding frequency in the fault table, and whether the ratio of the amplitude of the corresponding sideband component to the fundamental frequency component is within the ratio range.

[0103] Step S455: If the frequency meets the corresponding frequency in the fault table, and the ratio of the corresponding side frequency component to the fundamental frequency component amplitude is within the ratio range, it is determined that the asynchronous motor has an eccentricity fault and the inspection result is output.

[0104] In actual use, the increase in the number of pole pairs will lead to reduced output power and efficiency, increased noise and vibration, and reduced response speed. When selecting a motor, the appropriate number of pole pairs should be selected according to actual needs to obtain the best performance and effect. The motor speed and the number of pole pairs have the following quantitative relationship: n = 60f / p, that is, the 2-pole synchronous speed is 3000r / min, the 4-pole synchronous speed is 1500r / min, the 6-pole synchronous speed is 1000r / min, and the 8-pole synchronous speed is 750r / min. Due to the existence of slip, the actual motor speed is slightly lower than the speed of the rotating magnetic field. In practical applications, the rated speed of the motor is generally not less than 750r / min.

[0105] Therefore, the following classification discussion is made here:

[0106] (1) When p = 1, we get Table 1:

[0107]

[0108]

[0109] (2) When p = 2, we get Table 2:

[0110]

[0111] (3) When p=3, we get Table 3:

[0112]

[0113] (4) When p=4, we get Table 4:

[0114]

[0115]

[0116] Table 4

[0117] According to national standards GB / T24337-2009 and GBT 755 2019 / IEC 60034-1:2017 and experimental data, the ratio of the amplitude of the side frequency component to the fundamental frequency component should be within the range of 0.0024-0.02. Start the interharmonic signal generator to simulate the characteristic signal waveform of the eccentricity fault under different pole pairs, and use the tested product to collect and analyze the current signal. When the corresponding frequencies in Table 1, Table 2, Table 3, and Table 4 are met respectively, and the ratio of the amplitude to the fundamental frequency amplitude is within the range of 0.0024-0.02, it is determined that an eccentricity fault occurs.

[0118] If the inspection result shows that the asynchronous motor has an eccentric fault, the monitoring image of the asynchronous motor during operation is obtained, and the vibration data of the asynchronous motor is extracted by combining the attention mechanism, which is specifically as follows:

[0119] Step S51, extracting fault vibration data of the eccentricity fault from the historical fault data of the asynchronous motor eccentricity fault, and marking the eccentricity position of the fault vibration data.

[0120] Step S52, obtaining normal vibration data of the asynchronous motor when it works normally, and putting the fault vibration data and the normal vibration data into the attention mechanism model training.

[0121] Step S53, obtaining a monitoring image of the asynchronous motor during operation, and obtaining real-time vibration data of various parts of the asynchronous motor based on the monitoring image analysis.

[0122] Step S54, input the real-time vibration data into the trained attention mechanism model, filter out the data associated with the asynchronous motor eccentricity fault through the attention mechanism and output the vibration data.

[0123] In actual use, when an asynchronous motor has an eccentric fault, there will be specific vibrations. The eccentric position can be obtained through vibration and the severity of the eccentricity can be analyzed, but multi-point detection is required. Detection through vibration sensors requires multiple vibration sensors on the one hand, and on the other hand, the vibration sensors need to be installed on the asynchronous motor, which interferes with the operation of the asynchronous motor. Therefore, by monitoring the vibration data of the asynchronous motor, the interference to the asynchronous motor can be reduced. In addition, the attention mechanism can focus more accurately on the fault situation, reduce the interference of invalid data, and improve the accuracy and speed of data analysis. For example, after an asynchronous motor has an eccentric fault, the vibration does not change at all positions. The position of the asynchronous eccentricity is different, and the location of the vibration change is also different. The attention mechanism can realize local detection of fault changes and complete the vibration detection work more efficiently and quickly.

[0124] The steps of obtaining monitoring images of the asynchronous motor when it is running and obtaining real-time vibration data of various parts of the asynchronous motor according to the monitoring images are specifically as follows:

[0125] Step S531, continuously collecting monitoring images of the asynchronous motor when it is running, and extracting characteristic information of the asynchronous motor in the image after preprocessing the monitoring images.

[0126] Step S532, obtaining the change of the characteristic information in the monitoring image sequence, and obtaining the vibration displacement and vibration speed according to the change, so as to output the real-time vibration data of each part of the asynchronous motor.

[0127] In actual application, the collected images are preprocessed, including denoising, contrast enhancement, binarization and other operations, to improve image quality and facilitate subsequent analysis. Image processing algorithms are used to extract feature information related to vibration in the image. These feature information include the contour of the vibration waveform, the shape and size of the vibration area, and the temperature change caused by vibration. Specific marking points are set or identified on the surface of the motor, and these marking points are tracked and located using image processing algorithms. The vibration displacement of the motor is obtained by calculating the position change of the marking point in the image, and the vibration speed is obtained by calculating the time of the vibration displacement.

[0128] The steps of obtaining the vibration modes of different parts of the asynchronous motor, confirming the eccentricity position and severity of the eccentricity based on the vibration data, and obtaining the fault results are as follows:

[0129] Step S61, obtaining the vibration position corresponding to the most relevant vibration data in the attention mechanism as the eccentric position of the asynchronous motor.

[0130] By analyzing the spectral characteristics of the vibration signal, specific frequency components related to the eccentricity fault are identified. These frequency components are usually related to the vibration mode of the rotor or stator of the motor, and their amplitudes are affected by the eccentricity position and degree. According to the specific frequency components in the spectral characteristics and their corresponding amplitude changes, combined with the structural dynamic characteristics of the motor, the establishment and solution process of the motor vibration model, including the use of simulation technologies such as finite element analysis (FEA) to simulate the vibration behavior of the motor, estimates the location of the eccentricity fault.

[0131] Step S62, obtaining a connection drawing of the asynchronous motor, and calculating the distance between the asynchronous eccentric position and each component of the asynchronous motor according to the connection drawing and the asynchronous eccentric position, and recording it as the eccentric distance.

[0132] Step S63, obtaining the influence distance of the eccentricity fault on other components, searching for components of the asynchronous motor whose eccentricity distance is within the influence distance and recording them as the influence components.

[0133] Step S64, obtaining the severity of the eccentricity according to the resonance probability and functional importance of the influencing component.

[0134] Step S65: output the fault result in combination with the eccentricity position and the eccentricity severity.

[0135] In actual use, the unbalanced electromagnetic force generated by the eccentricity fault will act on the motor system as an excitation source, exciting the vibration mode of each component. When the excitation frequency is close to the natural frequency of a component or the entire system, resonance will be induced. Once resonance occurs, the vibration amplitude of the motor will increase significantly. This strong vibration will not only further aggravate the degree of eccentricity fault, such as causing increased bearing wear and rotor bending deformation. At the same time, resonance will also make the operating conditions of the motor worse, thereby accelerating the damage process of the motor. Therefore, the occurrence of resonance indicates that the eccentricity fault of the asynchronous motor is relatively serious and needs to be repaired in time. If the eccentric position is close to a component, then the component will be subjected to greater local stress under the action of the unbalanced electromagnetic force generated by the eccentricity fault, and the modal coupling effect between the component and its surrounding structure will be enhanced, and the stiffness of the component and its surrounding structure will be greatly affected, all of which will increase the risk of resonance. Therefore, according to the distance between the eccentric position and the component, the component that is prone to resonance is selected to analyze the resonance situation and judge the severity of the eccentricity.

[0136] According to the resonance probability and functional importance of the affected components, the steps to obtain the severity of eccentricity are as follows:

[0137] Step S641, obtaining the component vibration mode of the influencing component, obtaining the eccentric vibration mode of the asynchronous motor eccentric fault, comparing the component vibration mode and the eccentric vibration mode, and obtaining the modal similarity.

[0138] Step S642, obtaining the standard similarity of the modal resonance, screening the influencing components whose modal similarity reaches the standard similarity and recording them as faulty components.

[0139] Step S643, obtaining the importance degree BG of the function of the faulty component, and obtaining the influence degree BY of the faulty component on the asynchronous motor.

[0140] Step S644, obtaining the modal similarity BX corresponding to the faulty component, and calculating the eccentricity severity BP of the faulty component according to the eccentricity severity correlation function BP=b1×BG+b2×BY+b3×BX, where b1, b2, and b3 are proportional factors and are greater than 0.

[0141] In actual use, the closer the resonance mode of a component is to the fault resonance mode, the easier it is to resonate, that is, the higher the probability of resonance. The more important the function of the component is and the greater the impact on the asynchronous motor, the more serious the eccentricity fault is and the more timely maintenance is required. For example, the similarity between the resonance mode of component A and the fault resonance mode reaches 80%, and the function of component A is very important, which is the heat dissipation function. The heat dissipation function involves dissipating heat for multiple components, so it has a great impact on the operation of the asynchronous motor, indicating that the consequences of the eccentricity fault are more serious and the severity of the eccentricity fault is greater. The resonance similarity of component B is 20%, and it is an auxiliary independent function that does not affect the operation of other components of the asynchronous motor, so the severity is relatively small.

[0142] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for diagnosing eccentricity fault of an asynchronous motor, characterized in that: The following steps are involved: Acquire historical fault data of asynchronous motor eccentricity fault, and extract eccentricity fault features from the historical fault data; Diagnose whether the asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics, and output primary fault data if an eccentricity fault occurs; Obtain the current data of the asynchronous motor and the corresponding detection environment, perform quality screening on the current data according to the detection environment, and form current spectrum data; According to the current spectrum data, the frequency and amplitude in the current spectrum data are extracted, and according to the frequency and amplitude, whether the asynchronous motor has an eccentricity fault is detected and the detection result is output; If the inspection result shows that the asynchronous motor has an eccentric fault, the monitoring image of the asynchronous motor during operation is obtained, and the vibration data of the asynchronous motor is extracted by combining the attention mechanism; The vibration modes of different parts of the asynchronous motor are obtained, and the eccentricity position and severity are confirmed by combining the vibration data to obtain the fault results.

2. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 1, characterized in that: The step of diagnosing whether the asynchronous motor has an eccentricity fault according to the eccentricity fault characteristics and outputting primary fault data if an eccentricity fault occurs is specifically as follows: Statistically calculate the fault frequency AL of different eccentricity fault characteristics in the historical fault data of asynchronous eccentricity faults; Obtain historical fault data of all fault types of asynchronous motors, and count the number of fault types AS covered by the eccentricity fault characteristics; According to the fault eccentricity formula The fault eccentricity AP is calculated as follows: , is the scale factor and is greater than 0; Find the fault eccentricity corresponding to the eccentricity fault feature, superimpose the fault eccentricity of all extracted eccentricity fault features, and obtain the eccentricity fault probability; An eccentricity fault probability threshold is set to determine whether the eccentricity fault probability reaches the eccentricity fault probability threshold. If so, the asynchronous motor is diagnosed to have an eccentricity fault and primary fault data is output.

3. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 2, characterized in that: The step of obtaining the current data of the asynchronous motor and the corresponding detection environment, and performing quality screening on the current data according to the detection environment to form the current spectrum data is specifically as follows: Acquire the current data of the asynchronous motor and the detection environment during the current data detection to obtain real-time detection data; Obtain standard use environment data of the current detection device, compare and obtain the difference between the real-time detection data and the standard use environment data and record it as the environment difference; Set an environmental difference threshold, remove current data whose environmental difference reaches the environmental difference threshold, and record the removal time point of the removed current data; Set a time period, intercept the current data in the time period before the elimination time point, form a correlation function according to the change of the current data, and calculate the estimated current data at the elimination time point according to the correlation function; The predicted current data replaces the eliminated current data to form new current data, and the parameters of the current data are extracted to form current spectrum data.

4. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 3, characterized in that: The steps of extracting the frequency and amplitude in the current spectrum data according to the current spectrum data, and inspecting whether the asynchronous motor has an eccentricity fault according to the frequency and amplitude and outputting the inspection result are specifically as follows: According to the correlation function between motor speed and pole pair number Get the actual running speed of the asynchronous motor; Get the actual running speed n and synchronous speed of the motor and , according to the slip formula The slip rate s is calculated; Obtain relevant parameters of the motor, extract the main parameters affecting the current spectrum, and obtain the empirical formula for air gap eccentricity ,in is the power supply frequency, s is the slip rate, p is the number of magnetic pole pairs, and k is an arbitrary integer; The electromagnetic torque is calculated according to the torque formula ,in is the number of stator phases, p is the number of magnetic pole pairs, is the stator phase voltage, is the power supply frequency, and is the resistance and leakage reactance of each phase winding of the stator, and The rotor resistance and leakage reactance converted to the stator side; Obtain the standard fault range corresponding to the frequency and amplitude according to the actual operating speed, check whether the asynchronous motor has an eccentricity fault and output the test result; The steps of obtaining the standard fault range corresponding to the frequency and amplitude according to the actual operating speed, checking whether the asynchronous motor has an eccentricity fault and outputting the test result are specifically: Generate a mechanical characteristic curve according to the electromagnetic torque, obtain the value range of the slip rate s, and obtain the maximum and minimum values ​​of the slip rate s; Get the value of p, and discuss the different values ​​of p by category. When the value of p is fixed, the maximum value of slip rate s, the minimum value of slip rate s, and the value of k are used as variables for cross analysis to form a fault table. According to the frequency in the fault table, the ratio of the amplitude of the side frequency component to the fundamental frequency component corresponding to the frequency is calculated, and the ratio range of the amplitude of the side frequency component to the fundamental frequency component is obtained; Extract the frequency and amplitude from the current spectrum data, determine whether the frequency meets the corresponding frequency in the fault table, and whether the ratio of the corresponding side frequency component to the fundamental frequency component amplitude is within the ratio range; If the frequency meets the corresponding frequency in the fault table, and the ratio of the corresponding side frequency component to the fundamental frequency component amplitude is within the ratio range, it is determined that the asynchronous motor has an eccentricity fault and the test result is output.

5. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 4, characterized in that: If the inspection result is that the asynchronous motor has an eccentric fault, the monitoring image of the asynchronous motor when running is obtained, and the vibration data of the asynchronous motor is extracted by combining the attention mechanism, which is specifically as follows: Extracting fault vibration data of eccentricity fault from historical fault data of asynchronous motor eccentricity fault, and marking the eccentricity position of fault vibration data; Obtain the normal vibration data of the asynchronous motor when it is working normally, and put the fault vibration data and normal vibration data into the attention mechanism model training; Obtain monitoring images of the asynchronous motor during operation, and obtain real-time vibration data of various parts of the asynchronous motor based on the monitoring image analysis; The real-time vibration data is input into the trained attention mechanism model, and the data associated with the asynchronous motor eccentricity fault is filtered out through the attention mechanism and the vibration data is output.

6. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 5, characterized in that: The step of acquiring the monitoring image of the asynchronous motor when it is running and obtaining the real-time vibration data of each part of the asynchronous motor according to the monitoring image analysis is specifically as follows: Continuously collect monitoring images of the asynchronous motor when it is running, and extract characteristic information of the asynchronous motor in the image after preprocessing the monitoring images; The changes of characteristic information in the monitoring image sequence are obtained, and the vibration displacement and vibration velocity are obtained according to the changes, so as to output the real-time vibration data of each part of the asynchronous motor.

7. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 6, characterized in that: The steps of obtaining the vibration modes of different components of the asynchronous motor, confirming the eccentricity position and the severity of the eccentricity in combination with the vibration data, and obtaining the fault result are specifically as follows: Obtain the vibration position corresponding to the most relevant vibration data in the attention mechanism as the eccentric position of the asynchronous motor; Obtain a connection drawing of the asynchronous motor, and calculate the distance between the asynchronous eccentric position and each component of the asynchronous motor according to the connection drawing and the asynchronous eccentric position and record it as the eccentric distance; Obtain the impact distance of the eccentricity fault on other components, find the components of the asynchronous motor whose eccentricity distance is within the impact distance and record them as the impact components; The severity of eccentricity is obtained based on the resonance probability and functional importance of the affected components; The fault result is output based on the eccentricity position and eccentricity severity.

8. The inspection method for diagnosing eccentricity fault of an asynchronous motor according to claim 7, characterized in that: The step of obtaining the severity of eccentricity according to the resonance probability and functional importance of the influencing component is specifically as follows: Obtain the component vibration mode of the affected component, obtain the eccentric vibration mode of the asynchronous motor eccentricity fault, compare the component vibration mode and the eccentric vibration mode, and obtain the modal similarity; Obtain the standard similarity of modal resonance, select the affected components whose modal similarity reaches the standard similarity and record them as faulty components; Obtain the importance of the function of the faulty component BG, and obtain the influence of the faulty component on the asynchronous motor BY; Obtain the modal similarity BX corresponding to the faulty component and associate the function with the severity of eccentricity The eccentricity severity BP of the faulty component is calculated, where , , is the scaling factor and is greater than 0.

Images