Motor rotor eccentricity fault early warning device

Abstract

A motor rotor eccentricity fault early warning device comprises a detection unit used for collecting motor operation data in an operation state and comparing the motor operation data with motor data in an eccentricity fault state; the storage unit is used for storing data, eccentric faults and alarm data of the motor in a running state; and the early warning unit is used for timely notifying related maintenance personnel by generating early warning information and triggering an alarm mechanism through a detection unit algorithm. In the early warning generation process, the fault level is judged according to an abnormal detection result, and a notification is sent when a set threshold value is exceeded. Abnormal signals and characteristics thereof are recorded in a database for subsequent analysis. And the display unit is used for displaying the operation data of the motor in real time and displaying early warning information when an early warning prompt is triggered. According to the motor rotor eccentricity fault early warning device provided by the invention, accurate early warning can be carried out on the dynamic eccentricity fault of the motor.

Description

Technical Field

[0001] The invention relates to the technical field of motor fault diagnosis, and in particular to a motor rotor eccentricity fault early warning device. Background Art

[0002] Motors are widely used in people's production and life, and ensuring their safe and smooth operation is of great significance. The eccentricity fault of the motor is a fault caused by the uneven air gap between the stator and the rotor. Due to the influence of processing and assembly errors and loads during operation, the motor will more or less have the problem of uneven air gap between the stator and the rotor.

[0003] Rotor eccentricity is a common fault that accounts for a large proportion of motor failures. Currently, there are many solutions for early warning and online detection of motor rotor eccentricity, but there is no effective experimental device to accurately warn of motor rotor eccentricity, making it difficult to conduct in-depth research on motor eccentricity, especially the research on early warning of motor dynamic eccentricity. Eccentricity will cause the electromagnetic force fluctuation of the motor to increase, generating vibration and noise; and the electromagnetic force fluctuation and vibration will in turn intensify the movement of the stator and rotor, worsening the eccentricity. Such iterations produce a vicious cycle, and in severe cases, even cause stator and rotor sweep accidents, resulting in motor damage. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a motor rotor eccentricity fault warning device, which can effectively warn of motor rotor eccentricity faults, especially accurately warn of motor dynamic eccentricity faults, thereby improving the accuracy of motor rotor eccentricity fault judgment.

[0005] The technical solution adopted by the present invention is: A motor rotor eccentricity fault early warning device comprises the following units: Detection unit: collects the motor operation data in the running state, compares it with the motor data in the eccentricity fault state, and determines whether an eccentricity fault occurs; Storage unit: stores data on the motor's operating status, eccentricity faults and alarm data, as well as motor current, voltage, speed, acceleration, and continuous operating time; Early warning unit: The detection unit algorithm generates early warning information and triggers the alarm mechanism to promptly notify relevant maintenance personnel. The early warning generation process determines the fault level based on the results of abnormal detection and sends a notification when it exceeds the set threshold. The abnormal signal and its characteristics will be recorded in the database for subsequent analysis. In addition, the algorithm will trigger sound and light alarm devices such as buzzers and central display screens to improve the response speed of maintenance personnel; Display unit: Displays motor operation data in real time and displays warning information when a warning prompt is triggered.

[0006] The implementation method of the detection unit mainly includes the following steps: S1: Preprocess the collected motor operation signal, including filtering, denoising and normalization, to improve the signal quality. Use Butterworth filter to filter the signal to remove high-frequency noise; then use discrete wavelet transform to remove interference components in the signal; finally, perform minimum-maximum normalization to adjust the signal to a uniform range to ensure the stability of subsequent processing; S2: By analyzing the time domain, frequency domain and time-frequency domain features of the signal, key information reflecting the motor operating status is extracted. The time domain features include the mean, variance and other statistical quantities of the signal, the frequency domain features obtain the frequency components of the signal through fast Fourier transform, and the time-frequency domain features use short-time Fourier transform to analyze the time-frequency distribution of the signal; S3: Based on the extracted feature data, a fault detection model is constructed. The model is trained using a support vector machine, the model parameters are optimized through cross-validation, and the accuracy of the model is finally verified through a test set to ensure accurate identification of eccentricity faults; S4: Use the isolation forest algorithm to detect abnormalities in the current operating state of the motor. By calculating the difference between the characteristic value and the baseline characteristic value of the normal state, determine whether there is an abnormality, and set a reasonable threshold for fault warning.

[0007] Furthermore, the detection unit includes a signal sensor, which is installed on the motor rotor shaft. The signal sensor is respectively connected to the display unit, the early warning unit and the storage unit. The signal sensor senses signals of the motor rotor shaft speed and eccentricity, and realizes eccentricity fault early warning through signal preprocessing, feature extraction, model training and prediction, and anomaly detection.

[0008] The signal sensor is arranged on the motor rotor shaft via a bracket, the bracket is fixed to the motor rotor shaft, and the signal sensor starts to collect motor signals during the process of the motor rotor shaft rotating from stationary to rotating.

[0009] A transceiver module is arranged between the detection unit and the early warning unit, and the transceiver module includes a wirelessly connected signal transmitter and a signal receiver. The signal transmitter is fixedly mounted on the rotating shaft, and the signal receiver connects the early warning unit and the storage unit. The signal transmitter sends the signal of the detection unit to the signal receiver, and the signal receiver transmits the received signal to the storage unit for storage. The early warning unit also includes a buzzer. If the signal detected by the detection unit differs from the benchmark operating experience characteristic value, a sound and a visual alarm will be issued through the buzzer and the central display screen respectively.

[0010] The present invention proposes a motor rotor eccentricity fault early warning device, and the technical effects are as follows: 1) The present invention is conducive to motor maintenance personnel to inspect and repair key equipment in advance, which can effectively avoid failure of key equipment and affect normal operation.

[0011] 2) The present invention collects and stores motor data, thereby increasing monitoring of abnormal data and continuously improving the accuracy of fault judgment.

[0012] 3) The present invention can automatically detect the working condition of the motor in real time, is simple to maintain, and can store data. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Schematic diagram of dynamic eccentricity principle.

[0014] Figure 2 This is a schematic diagram of the static eccentricity principle.

[0015] Figure 3 This is a schematic diagram of the front structure of the motor detection unit of this embodiment.

[0016] Figure 4 It is a schematic diagram of the back structure of the motor detection unit of this embodiment.

[0017] Figure 5 Schematic diagram of the early warning unit of this embodiment.

[0018] Figure 6 It is a schematic diagram of the display unit and the storage unit of this embodiment.

[0019] In the figure: signal sensor 1, motor stator 2, vent 3, motor rotor 4, motor housing 5, signal transmitter 6, motor rotor shaft 7. DETAILED DESCRIPTION

[0020] like Figure 1 , Figure 2 As shown in the figure, the air gap eccentricity of the motor is a phenomenon in which the radial air gap length between the stator a and the rotor b is uneven, which exists in almost all motors to a greater or lesser extent. According to the uneven distribution characteristics, the air gap eccentricity can be divided into two basic types: axial uniform eccentricity (static eccentricity, dynamic eccentricity) and axial uneven eccentricity (symmetrical oblique eccentricity).

[0021] like Figure 1 As shown in the figure, the stator a and the rotor b are not concentric, the rotor b takes its own geometric axis as the axis, and the position of the minimum air gap d remains unchanged, which belongs to the uniform eccentricity of the air gap. When there is an eccentric distance e between the center of the rotor b and the center of the stator a, and the minimum air gap d between the rotor b and the stator a remains unchanged, it indicates that the motor has static eccentricity.

[0022] like Figure 2As shown in the figure, the stator a and the rotor b are not concentric, but the rotor b is centered on the geometric axis of the stator a, and the position of the minimum air gap d changes dynamically, which is an uneven air gap eccentricity. When there is an eccentric distance e between the center of the rotor b and the center of the stator a, and the minimum air gap d between the rotor b and the stator a keeps changing, it indicates that the motor has dynamic eccentricity.

[0023] like Figure 3~Figure 6 As shown, a motor rotor eccentricity fault warning device includes a detection unit, a display unit, a warning unit and a storage unit. The detection unit includes a signal sensor 1, which is installed on the motor rotor shaft 7. Under the rotation of centrifugal force, the motor rotor 4 applies a certain pressure to the signal sensor 1. The centrifugal force on the signal sensor 1 is related to the speed and eccentricity of the motor rotor shaft 7.

[0024] The magnitude of centrifugal force is related to the mass of the object, the radius of rotation and the angular velocity. The calculation formula is:

[0025] in represents the centrifugal force, represents the mass of an object, represents the radius of rotation, It indicates angular velocity. From the formula, we can see that the centrifugal force is proportional to the mass of the object, proportional to the square of the rotation radius, and proportional to the square of the angular velocity.

[0026] Speed ​​refers to the angle of rotation per unit time, usually expressed in revolutions per minute. The calculation formula is:

[0027] in represents the angular velocity, It represents the number of revolutions per minute, and 60 represents one minute. It represents the circumference of a circle. From the formula, we can see that the angular velocity is proportional to the number of revolutions per minute.

[0028] Through the calculation formula of centrifugal force and the calculation formula of speed, we can get the calculation formula between centrifugal force and speed. Substituting the calculation formula of speed into the calculation formula of centrifugal force, we get:

[0029] Through this formula, the magnitude of the centrifugal force can be calculated based on the mass of the object, the rotation radius and the number of revolutions per minute. This formula is applicable to uniform circular motion, and the rotor eccentricity factor must also be considered when the motor fails.

[0030] The signal sensor 1 is arranged on the motor rotor shaft 7 through a bracket, and the bracket is fixed to the motor rotor shaft 7. When the motor rotor shaft 7 rotates from stationary to rotating, the signal sensor 1 starts to collect motor signals.

[0031] In order not to affect the rotation of the motor rotor shaft 7, the bracket adopts a central symmetrical circular structure. Figure 3 and Figure 4 The motor includes a motor stator 2 and a motor rotor 4 coaxial with the motor rotor shaft 7. In order to reduce the mass of the rotating device, the outer ring of the motor rotor shaft 7 is provided with ventilation holes, as shown in FIG. Figure 3 The vents 3 in the motor housing can reduce the weight of the rotating equipment, reduce energy loss and the risk of failure. At the same time, the motor housing 5, as an important part of the motor, is mainly used to protect the components inside the motor from the influence of the external environment, while shielding the electromagnetic interference from the outside world to ensure the normal operation of the motor in a complex environment. It is made of cast aluminum alloy material. Cast aluminum alloy has excellent mechanical strength, corrosion resistance and good heat dissipation performance, and can effectively protect the precision parts inside the motor. In addition, the aluminum alloy material is light in weight, which helps to reduce the weight of the entire motor and improve energy efficiency.

[0032] In order to solve the influence of electric field and magnetic field on the components of the device, in this embodiment, the signal sensor 1 adopts ADXL345 acceleration sensor. A three-axis acceleration sensor has a measurement range of up to ±16g and a data output rate of up to 3200Hz. It can be mounted on the connecting rod after opening a hole, or the center of the connecting rod is opened and the rotating shaft is inserted into the center hole, such as Figure 3 As shown. The shaft is rotated through the connecting rod. When the motor rotor shaft 7 rotates, under the action of centrifugal force, the signal sensor 1 will be closer to the inside of the motor rotor, which can better collect signals. The vibration data is transmitted to the detection unit in real time. The main working steps of the detection unit are as follows: 1. Signal preprocessing In the detection unit, signal preprocessing is a key step to ensure the accuracy and consistency of the data obtained from the signal sensor 1. First, the signal sensor 1 is installed on the motor rotor shaft 7 to collect the signal data of the motor in real time during operation. In order to eliminate high-frequency noise and low-frequency interference in the signal, a Butterworth filter is used to filter the signal. Its transfer function is:

[0033] in, is the cut-off frequency, n is the order of the filter.

[0034] After filtering, the signal is further denoised by discrete wavelet transform (DWT), the main components of the signal are extracted, useless high-frequency noise is removed, and the signal is reconstructed. Finally, the minimum-maximum normalization method is used to scale the signal amplitude to a uniform range. The normalization formula is:

[0035] in and , are the minimum and maximum values ​​of the signal respectively. Through this series of preprocessing steps, the accuracy and consistency of the signal are effectively improved, laying the foundation for subsequent feature extraction and fault detection.

[0036] 2. Feature extraction After completing the signal preprocessing, the detection unit performs feature extraction. By analyzing the time domain, frequency domain and time-frequency domain characteristics of the signal, the characteristics of the motor rotor eccentricity fault can be effectively captured. The time domain characteristics include statistical quantities such as the mean, variance, kurtosis and skewness of the signal. For example, the calculation formula for the mean is:

[0037] Frequency domain feature extraction uses fast Fourier transform (FFT) to analyze the frequency components and power spectrum density of the signal. The conversion formula is:

[0038] In addition, the time-frequency domain feature extraction uses short-time Fourier transform (STFT) to obtain the local characteristics of the signal in time and frequency. The calculation formula is:

[0039] Through comprehensive analysis of these characteristics, the specific manifestations of motor rotor eccentricity fault can be accurately identified.

[0040] 3. Model training and prediction After feature extraction is completed, the detection unit inputs the extracted feature data into the fault detection model. The support vector machine (SVM) is selected as the main fault detection model, and its decision function is:

[0041] in, w is the weight vector, b By analyzing and labeling the historical data set, different fault types can be classified. In the model training process, the cross-validation method is used to optimize the model parameters, and the method of minimizing the loss function is used:

[0042] Finally, the accuracy of the model is verified through the test set, and performance indicators such as accuracy, precision, and recall are calculated to evaluate the generalization ability of the model.

[0043] 4. Anomaly Detection The detection unit's algorithm also includes anomaly detection, which identifies potential faults by comparing the current signal features with the baseline features of normal data. First, the difference between the current features and the normal features is calculated using the Euclidean distance metric:

[0044] in: x i is the current feature vector x No. i Quantity, μ i is a normal eigenvector μ No. i Quantity, n is the dimension of the feature vector.

[0045] In order to further detect anomalies, the isolation forest algorithm is used to detect anomalies in feature data. The isolation forest algorithm is based on a tree structure, and the calculation formula for the anomaly score is:

[0046] in E(h(x)) is the average path length between trees in an isolation forest. h(x) It is the path length from the root node to the leaf node. A longer path length usually indicates that the data points are sparse, that is, anomalies. Set a reasonable threshold based on historical data to determine whether there is an anomaly. You can set a percentile threshold based on the distribution of the anomaly score, such as the 95th percentile. If the anomaly score exceeds the set threshold, it is judged as an anomaly. The threshold setting formula is as follows:

[0047] Represents the 95th percentile of the anomaly score S.

[0048] After detecting an abnormality, the detection unit algorithm will generate warning information according to the fault level and notify the warning unit. In order to facilitate communication between the detection unit and the warning unit, a transceiver module is set between the two. The transceiver module specifically includes a wirelessly connected signal transmitter 6 and a Figure 5The signal receiver 9 shown. The early warning unit uses the STM32F103C8T6 microcontroller as the core processor. The STM32F103C8T6 is based on the ARM Cortex-M3 core and has a main frequency of 72 MHz. The microcontroller receives and processes vibration data from the ADXL345 acceleration sensor. The signal transmitter 6 uses the NRF24L01 wireless communication module, which operates in the 2.4GHz ISM band, with a data rate of up to 2Mbps, a transmission distance of about 100 meters indoors, and a line of sight of up to 1000 meters. The signal receiver 9 also uses the NRF24L01 wireless communication module to receive wireless data from the signal transmitter and transmit it to a storage unit or other processing unit for analysis and storage. The high sensitivity and low power consumption design of the NRF24L01 ensures the stability and efficiency of data transmission. The signal transmitter 6 sends the signal of the detection unit to the signal receiver 9, and the signal receiver 9 receives and transmits it to the data storage unit for storage.

[0049] The storage unit uses SST26VF064B flash memory with a capacity of 64Mb, which is connected to the early warning unit. The flash memory is used to store motor operation data and fault logs, which is convenient for subsequent fault analysis and maintenance records. SST26VF064B has a data retention time of more than 100 years and more than 100,000 erase and write times, ensuring the reliability and durability of data storage.

[0050] like Figure 6 The display unit shown includes a central display screen 10, a slow forward key 11 and a fast forward key 15, a volume down key 12 and a volume up key 14, and a start / stop key 13. A microprocessor is integrated inside the display unit to process the data transmitted from the signal sensor and the storage unit, and to display the processed information on the screen in real time. The microprocessor has efficient data processing capabilities and can ensure the smooth operation of the display unit under high data volume. After receiving the signal, the data storage unit will sort it and play the sorted signal on the display. The central display screen 10 uses a common display.

[0051] The detection unit algorithm generates early warning information and triggers the alarm mechanism to promptly notify the relevant maintenance personnel. The early warning generation process determines the fault level based on the results of the abnormal detection and sends a notification when it exceeds the set threshold. The judgment result is displayed on the central display screen 10, and a signal is sent to the early warning unit for early warning reminder. When the single chip detects abnormal vibration, the early warning unit will trigger the alarm mechanism, including control Figure 5 Buzzer 8 and Figure 6 The central display screen 10 in the system will send out audible and visual alarms to remind maintenance personnel to carry out inspection and repair. Abnormal signals and their characteristics will be recorded in the data storage unit for subsequent analysis.

[0052] Through the above hardware configuration and implementation method, this embodiment can monitor the running state of the motor rotor in real time, timely discover and warn the motor rotor eccentricity fault, and effectively improve the safety and reliability of the motor operation. The implementation method of this embodiment can be flexibly adjusted and expanded according to actual needs. For example, different types of acceleration sensors, single-chip microcomputers and wireless communication modules can be selected to adapt to different application scenarios and needs.

Claims

1. A motor rotor eccentricity fault early warning device, characterized in that: The following units are included: Detection unit: collects the motor operation data in the running state, compares it with the motor data in the eccentricity fault state, and determines whether an eccentricity fault occurs; Storage unit: stores data on the motor's operating status, eccentricity faults and alarm data, as well as motor current, voltage, speed, acceleration, and continuous operating time; Early warning unit: The detection unit algorithm generates early warning information and triggers the alarm mechanism to promptly notify relevant maintenance personnel; The warning generation process determines the fault level based on the results of anomaly detection and sends a notification when the set threshold is exceeded; the abnormal signal and its characteristics will be recorded in the database for subsequent analysis; in addition, the algorithm will trigger the sound and light alarm equipment to improve the response speed of maintenance personnel; Display unit: Displays motor operation data in real time and displays warning information when a warning prompt is triggered.

2. The motor rotor eccentricity fault early warning device according to claim 1, characterized in that: The main steps of the implementation method of the detection unit include: S1: Preprocess the collected motor operation signal, including filtering, denoising and normalization. Use Butterworth filter to filter the signal to remove high-frequency noise. Then use discrete wavelet transform to remove interference components in the signal. Finally, perform minimum-maximum normalization to adjust the signal to a uniform range to ensure the stability of subsequent processing. S2: By analyzing the time domain, frequency domain and time-frequency domain features of the signal, key information reflecting the motor operating status is extracted; the time domain features include the mean and variance of the signal; the frequency domain features obtain the frequency components of the signal through fast Fourier transform, and the time-frequency domain features use short-time Fourier transform to analyze the time-frequency distribution of the signal; S3: Based on the extracted feature data, a fault detection model is constructed; the model is trained using a support vector machine, the model parameters are optimized through cross-validation, and finally the accuracy of the model is verified through a test set to ensure accurate identification of eccentricity faults; S4: Use the isolation forest algorithm to detect abnormalities in the current operating state of the motor. By calculating the difference between the characteristic value and the baseline characteristic value of the normal state, determine whether there is an abnormality, and set a reasonable threshold for fault warning.

3. The motor rotor eccentricity fault early warning device according to claim 2, characterized in that: The detection unit comprises a signal sensor (1), the signal sensor (1) being mounted on a motor rotor shaft (7), the signal sensor (1) being connected to a display unit, a warning unit and a storage unit respectively, the signal sensor (1) sensing signals of the motor rotor shaft (7) speed and eccentricity, and achieving eccentricity fault warning through signal preprocessing, feature extraction, model training and prediction, and abnormality detection.

4. The motor rotor eccentricity fault early warning device according to claim 3 is characterized in that: The signal sensor (1) is arranged on the motor rotor shaft (7) via a bracket, the bracket is fixed to the motor rotor shaft (7), and when the motor rotor shaft (7) rotates from stationary to rotating, the signal sensor (1) starts to collect motor signals.

5. The motor rotor eccentricity fault early warning device according to claim 4, characterized in that: A transceiver module is provided between the detection unit and the early warning unit, the transceiver module comprising a wirelessly connected signal transmitter (6) and a signal receiver (9), the signal transmitter (6) being fixedly mounted on the motor rotor shaft (7), and the signal receiver (9) being connected to the early warning unit and the storage unit; the signal transmitter (6) transmits a signal from the detection unit to the signal receiver (9), and the signal receiver (9) transmits the received signal to the storage unit for storage; the early warning unit further comprises a buzzer (8), and if the signal detected by the detection unit differs from a reference operating experience characteristic value, an audible and visual alarm is emitted through the buzzer (8) and the central display screen (10), respectively.

Images