A motor vibration suppression system, device and method based on active magnetic field compensation
The motor vibration suppression system with active magnetic field compensation uses a compensation module to generate a reverse compensation electromagnetic field. Combined with a machine learning model and an automatic controller, it solves the wear and passivity problems of existing motor vibration suppression schemes, realizes wear-free active vibration suppression and dynamic adjustment, and improves the motor control accuracy and stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN HUACHENG IND CONTROL
- Filing Date
- 2024-12-12
- Publication Date
- 2026-06-12
AI Technical Summary
Existing motor vibration suppression solutions rely on mechanical damping devices or damping materials, which are subject to wear risks and can only provide passive damping without dynamic adjustment, thus affecting the motor control accuracy and stability.
An active magnetic field compensation-based motor vibration suppression system is adopted. Vibration sensors monitor data, a compensation module generates a reverse compensation electromagnetic field, and machine learning models and automatic controllers are used for dynamic adjustment to generate control signals to suppress vibration.
It achieves wear-free active vibration suppression, has a wide range of applications, can be dynamically adjusted, improves motor control accuracy and stability, shortens trial and error time, and enhances system stability and reliability.
Smart Images
Figure CN119766045B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of motor technology, and in particular to a motor vibration suppression system, device, and method based on active magnetic field compensation. Background Technology
[0002] Permanent magnet synchronous servo motors are synchronous motors that use permanent magnets to generate magnetic fields. However, they suffer from significant vibration issues at high speeds, which not only affect motor control accuracy but also negatively impact equipment lifespan and operational stability. How to suppress motor vibration and thus improve motor control accuracy and stability is a technical problem that needs to be solved in the field of motors.
[0003] Existing motor vibration suppression solutions mainly rely on mechanical damping devices or damping materials. These devices or materials typically need to be in direct contact with the motor body, posing a risk of motor wear. Furthermore, existing motor vibration suppression solutions can only passively reduce vibration, usually only addressing vibrations within a specific frequency range, and cannot dynamically adjust according to changes in motor operating conditions. Summary of the Invention
[0004] Therefore, the purpose of this invention is to provide a motor vibration suppression system, device, and method based on active magnetic field compensation, which actively generates a compensating electromagnetic field based on motor vibration data. This not only eliminates the risk of motor wear but also provides better vibration reduction and a wider range of applications.
[0005] This invention provides a motor vibration suppression system based on active magnetic field compensation, comprising: a vibration sensor for monitoring motor vibration and collecting vibration data; a control device for generating a control signal based on the vibration data; and a compensation module for generating a compensation electromagnetic field based on the control signal, thereby generating a reverse compensation force on the motor based on the compensation electromagnetic field. The compensation module includes several sets of electromagnetic coils and a drive unit electrically connected to the electromagnetic coils. The drive unit adjusts the drive current on the electromagnetic coils according to the control signal, thereby controlling the electromagnetic coils to generate a specified compensation electromagnetic field.
[0006] This invention applies a counter-compensating force to the motor to suppress vibration by compensating for the electromagnetic field. This eliminates the need for direct contact with the motor body, thus avoiding the risk of motor wear and ensuring the motor's lifespan. As an active rather than passive vibration damping method, this invention offers superior vibration suppression. Furthermore, it can dynamically adjust to changes in the motor's operating conditions, effectively handling complex and varied vibration scenarios. Moreover, the compensating electromagnetic field generated by the compensation module is independent of the rotating magnetic field generated by the motor's original stator windings. This not only compensates for vibrations in directions other than motor rotation, broadening its applicability, but also allows for independent operation without interfering with the motor drive, improving system stability and reliability.
[0007] Furthermore, the control device includes a machine learning model, a feedback unit, and an automatic controller. The control device generates a control signal through the following process: the machine learning model generates an initial basic control signal based on historical vibration data; the feedback unit controls the compensation module to generate a compensation electromagnetic field based on the initial basic control signal, while simultaneously acquiring current vibration data through the vibration sensor; the machine learning model updates the basic control signal based on the historical vibration data and the current vibration data; and the automatic controller fine-tunes the basic control signal based on the current vibration data to obtain the final control signal.
[0008] Furthermore, the control device also includes a data update unit and a loop unit; after the automatic controller receives the control signal, the data update unit merges the current vibration data of the previous moment with the historical vibration data to obtain new historical vibration data, and acquires new current vibration data through the vibration sensor; the loop unit repeatedly calls the machine learning model, the automatic controller and the data update unit, thereby continuously updating the control signal according to the current vibration data.
[0009] Furthermore, the machine learning model is trained using historical vibration data, and the parameters of the machine learning model are continuously updated as the historical vibration data is updated.
[0010] Furthermore, the automatic controller specifically includes: a feature extraction unit, used to analyze the current vibration data and extract vibration feature values; a reverse compensation force calculation unit, used to calculate the corresponding reverse compensation force based on the vibration feature values; a compensation electromagnetic field calculation unit, used to calculate the corresponding compensation electromagnetic field based on the reverse compensation force; a fine-tuning amount generation unit, used to generate a corresponding control signal fine-tuning amount based on the compensation electromagnetic field; and a control signal output unit, used to superimpose the basic control signal and the control signal fine-tuning amount to obtain the fine-tuned control signal.
[0011] Furthermore, it also includes a magnetic field sensor for detecting the actual compensation electromagnetic field; the control device further includes a magnetic field correction unit; the magnetic field correction unit, after the fine-tuning unit obtains the control signal fine-tuning amount, compares the difference between the actual compensation electromagnetic field and the target compensation electromagnetic field obtained by the compensation electromagnetic field calculation unit, and corrects the control signal fine-tuning amount.
[0012] Furthermore, the control device also includes a preprocessing unit; after acquiring the current vibration data, the preprocessing unit performs normalization and filtering noise reduction on the current vibration data, and then outputs the preprocessed current vibration data to the machine learning model and the automatic controller.
[0013] Furthermore, the machine learning model is an LSTM model.
[0014] Based on the same inventive concept, the present invention also provides a control method for the above-mentioned motor vibration suppression system, which includes the following steps: S1, generating an initial basic control signal based on historical vibration data; S2, generating a compensation electromagnetic field based on the initial basic control signal, and simultaneously acquiring current vibration data; S3, updating the basic control signal based on historical vibration data and current vibration data; S4, fine-tuning the basic control signal based on the current vibration data to obtain a control signal.
[0015] Based on the same inventive concept, the present invention also provides a control device applied to the above-mentioned motor vibration suppression system, comprising: a machine learning model, a feedback unit, and an automatic controller; the machine learning model generates an initial basic control signal based on historical vibration data; the feedback unit controls the compensation module to generate a compensation electromagnetic field based on the initial basic control signal, and simultaneously acquires current vibration data through the vibration sensor; the machine learning model updates the basic control signal based on the historical vibration data and the current vibration data; the automatic controller fine-tunes the basic control signal based on the current vibration data to obtain a control signal.
[0016] To better understand and implement this invention, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of a motor vibration suppression system based on active magnetic field compensation according to an embodiment of the present invention.
[0018] Figure 2 This is a schematic diagram showing the installation positions of the vibration sensor and compensation module according to an embodiment of the present invention;
[0019] Figure 3 This is a schematic diagram of a control device according to an embodiment of the present invention;
[0020] Figure 4 This is a schematic flowchart of a control method executed by a control device according to an embodiment of the present invention. Detailed Implementation
[0021] To address the shortcomings of existing mechanical vibration damping devices or damping materials, which can wear down the motor and only provide passive vibration reduction, this invention adds a compensation module at the motor bearing or rotor to generate a compensating electromagnetic field. By controlling this compensating electromagnetic field, a reverse compensating force is applied to the motor to suppress vibration. This invention does not require direct contact with the motor body and can actively compensate for vibrations in any direction and mode, exhibiting good vibration suppression effect and a wide range of applications.
[0022] Please see Figure 1 , Figure 1 This is a schematic diagram of a motor vibration suppression system based on active magnetic field compensation according to an embodiment of the present invention. The motor vibration suppression system based on active magnetic field compensation of the present invention includes: a vibration sensor 1, a control device 2, and a compensation module 3. The vibration sensor 1 monitors the motor vibration and collects vibration data. The control device executes the motor vibration suppression control method of the present invention based on the vibration data, generating a control signal. The compensation module generates an adjustable compensation electromagnetic field based on the control signal, generating a reverse compensation force on the motor based on the compensation electromagnetic field; the reverse compensation force is opposite to the vibration direction of the motor, thereby suppressing motor vibration.
[0023] Please see Figure 2 , Figure 2 This is a schematic diagram showing the installation position of the vibration sensor and compensation module according to an embodiment of the present invention.
[0024] Specifically, the vibration sensor 1 includes: an acceleration sensor, a displacement sensor, a velocity sensor, and / or a position sensor mounted on the motor bearing or rotor. The vibration data includes: acceleration data, displacement data, velocity data, and / or position data. Preferably, there are multiple vibration sensors, each disposed at different measurement points on the motor bearing or rotor, thereby collecting vibration data from multiple angles.
[0025] Specifically, the compensation module 3 includes: several sets of electromagnetic coils disposed on the motor bearings or rotor, and a drive unit electrically connected to the electromagnetic coils. The drive unit adjusts the drive current output to the electromagnetic coils according to the control signal output by the control device, thereby controlling the electromagnetic coils to generate a specified compensation electromagnetic field.
[0026] Please see Figure 3-4 , Figure 3 This is a schematic diagram of the control device according to an embodiment of the present invention. Figure 4This is a schematic flowchart of a control method executed by a control device according to an embodiment of the present invention.
[0027] Specifically, the control device 2 includes: a machine learning model, a feedback unit, an automatic controller, a data update unit, and a loop unit.
[0028] The machine learning model executes step S1: generating an initial basic control signal based on historical vibration data. Step S1 can be completed before the motor starts running, so that the initial basic control signal can be obtained immediately once the motor starts running. This avoids overshoot problems caused by lack of data in the initial stage of motor operation, greatly shortens the trial and error time, and improves control efficiency.
[0029] The feedback unit executes step S2: controlling the compensation module 3 to generate a compensation electromagnetic field according to the initial basic control signal, and simultaneously acquiring the current vibration data through the vibration sensor 1.
[0030] The machine learning model executes step S3: updating the basic control signal based on historical vibration data and current vibration data.
[0031] The automatic controller executes step S4: fine-tuning the basic control signal based on the current vibration data to obtain a control signal.
[0032] The data update unit executes step S5: controlling the compensation module 3 to generate a compensation electromagnetic field according to the control signal; merging the current vibration data of the previous moment with the historical vibration data to obtain new historical vibration data; and acquiring new current vibration data through the vibration sensor 1.
[0033] The loop unit executes step S6: repeatedly calling the machine learning model, automatic controller and data update unit to repeat steps S3-S5, thereby continuously updating the control signal according to the current vibration data until the motor stops running and the repetition of steps S3-S5 ends.
[0034] The machine learning model is used to extract and map features from historical and current vibration data to generate a basic control signal, thus providing a fundamental reference for generating the final control signal. Compared to directly generating the final control signal, this invention first obtains a macroscopic basic control signal and then fine-tunes it, combining macroscopic and microscopic adjustments. This not only improves the accuracy of the final generated control signal but also shortens the trial-and-error time, achieving faster computation speed and higher cost-effectiveness.
[0035] The machine learning model is trained from historical vibration data. When the data update unit updates the historical vibration data, the parameters of the machine learning model are also continuously updated, thereby enabling it to adapt to complex and ever-changing motor operating conditions.
[0036] The machine learning model can be a neural network model capable of learning historical data features, such as RNN (Recurrent Neural Network), LSTM (Long Short-Term Memory), GRU (Gate Recurrent Unit), or Bi-LSTM (Bidirectional Long Short-Term Memory). In this embodiment, the machine learning model is preferably an LSTM model.
[0037] The automatic controller specifically includes: a feature extraction unit, a reverse compensation force calculation unit, a compensation electromagnetic field calculation unit, a fine-tuning quantity generation unit, and a control signal output unit.
[0038] The feature extraction unit is used to perform step S51: analyze the current vibration data and extract vibration feature values.
[0039] The vibration characteristic values include, but are not limited to, vibration amplitude, vibration frequency, and vibration direction. Step S51 specifically involves: performing a fast Fourier transform on the current vibration data to obtain the vibration amplitude and vibration frequency; and performing mechanical analysis on the current vibration data (collected by multiple vibration sensors 1) at multiple points to obtain the vibration direction.
[0040] The reverse compensation force calculation unit is used to perform step S52: calculate the corresponding reverse compensation force based on the vibration characteristic value.
[0041] The compensation electromagnetic field calculation unit is used to execute step S53: calculate the corresponding compensation electromagnetic field based on the reverse compensation force.
[0042] The fine-tuning unit is used to execute step S54: generate a corresponding control signal fine-tuning amount based on the compensated electromagnetic field.
[0043] The control signal output unit is used to execute step S55: superimpose the basic control signal and the control signal fine-tuning amount to obtain the fine-tuned control signal.
[0044] Step S52 can be implemented using automatic control algorithms such as PID, MPC, SMC, adaptive control, or fuzzy control, with the control objective of minimizing the vibration characteristic value, to determine the magnitude and direction of the reverse compensation force. Given the relevant parameters of the electromagnetic coils of the motor and compensation module, a definite mapping relationship exists between the reverse compensation force, the compensation electromagnetic field, and the control signal. This mapping relationship can be determined using existing measurement techniques. Then, through steps S53-S54, the corresponding control signal fine-tuning amount is derived and calculated based on the reverse compensation force.
[0045] Furthermore, the motor vibration suppression system of the present invention also includes a magnetic field sensor. The magnetic field sensor is disposed at the motor bearing or rotor and is used to detect the actual compensation electromagnetic field. The control device of the present invention also includes a magnetic field correction unit. The magnetic field correction unit is used to perform the following step after the fine-tuning amount generation unit executes step S54: comparing the difference between the actual compensation electromagnetic field and the target compensation electromagnetic field calculated in step S53, and correcting the fine-tuning amount of the control signal.
[0046] Furthermore, the motor vibration suppression system of the present invention also includes a power management module. The power management module is used to dynamically adjust the power supply distribution of each module in the system; when the motor is running at high speed, priority is given to ensuring the power supply to the electromagnetic coil of the compensation module, so as to ensure that the strength of the reverse compensation force for suppressing motor vibration is sufficiently large.
[0047] The present invention has the following technical effects:
[0048] 1. This invention applies a counter-compensating force to the motor to suppress vibration by compensating for the electromagnetic field. It does not require direct contact with the motor body, eliminating the risk of motor wear and ensuring the motor's lifespan. This invention is an active vibration damping system, not a passive one, and offers better vibration suppression compared to passive damping. Furthermore, this invention can dynamically adjust to changes in the motor's operating conditions, enabling it to handle complex and varied vibration situations.
[0049] 2. Existing motor vibration suppression schemes based on magnetic field compensation typically achieve this by controlling the rotating magnetic field generated by the original stator windings of the motor. However, the rotating magnetic field generated by the original stator windings can only generate a reverse compensating force in the direction of motor rotation and cannot suppress vibrations in directions other than motor rotation, thus limiting its applicability. In contrast, the compensation module of this invention generates a compensation electromagnetic field independent of the rotating magnetic field generated by the original stator windings. This not only compensates for vibrations in directions other than motor rotation, thus expanding its applicability, but also allows it to operate independently without interfering with motor drive, improving system stability and reliability. It can respond to vibrations more quickly and achieve better vibration suppression results.
[0050] 3. The control method of this invention combines the advantages of machine learning models and traditional automatic control algorithms. First, the basic macroscopic control signal is obtained through the machine learning model, and then the basic control signal is fine-tuned through the traditional automatic control algorithm. This combines macroscopic and microscopic adjustment, which not only improves the accuracy of the final generated control signal, but also achieves a faster calculation speed and higher cost performance.
[0051] 4. The machine learning model of this invention can learn the characteristics of historical vibration data and obtain the initial basic control signal in advance, thereby avoiding the overshoot problem caused by lack of data in the initial stage of motor operation, greatly shortening the trial and error time, and improving the efficiency of the control algorithm. The machine learning model of this invention can also continuously update its own parameters based on updated historical vibration data, thereby adapting to complex and ever-changing motor operating conditions.
[0052] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and the present invention also intends to include these modifications and variations.
Claims
1. A motor vibration suppression system based on active magnetic field compensation, characterized in that, include: Vibration sensors monitor motor vibration and collect vibration data. The control device includes a machine learning model, a feedback unit, an automatic controller, a data update unit, and a loop unit; The control device generates control signals through the following process: The machine learning model generates initial basic control signals based on historical vibration data; The feedback unit controls the compensation module to generate a compensation electromagnetic field according to the initial basic control signal, and at the same time acquires the current vibration data through the vibration sensor. The machine learning model updates the basic control signal based on historical vibration data and current vibration data; The automatic controller fine-tunes the basic control signal based on the current vibration data to obtain a control signal; After the automatic controller receives the control signal, the data update unit merges the current vibration data from the previous moment with the historical vibration data to obtain new historical vibration data, and acquires new current vibration data through the vibration sensor. The loop unit repeatedly calls the machine learning model, the automatic controller, and the data update unit to continuously update the control signal based on the current vibration data. The compensation module generates a compensation electromagnetic field based on the control signal, and generates a reverse compensation force on the motor based on the compensation electromagnetic field. The compensation module includes several sets of electromagnetic coils and a drive unit electrically connected to the electromagnetic coils; the drive unit adjusts the drive current on the electromagnetic coils according to the control signal, thereby controlling the electromagnetic coils to generate a specified compensation electromagnetic field.
2. The motor vibration suppression system based on active magnetic field compensation according to claim 1, characterized in that: The machine learning model is trained using historical vibration data, and its parameters are continuously updated as the historical vibration data is updated.
3. The motor vibration suppression system based on active magnetic field compensation according to claim 2, characterized in that: The automatic controller specifically includes: A feature extraction unit is used to analyze the current vibration data and extract vibration feature values; A reverse compensation force calculation unit is used to calculate the corresponding reverse compensation force based on the vibration characteristic value; A compensation electromagnetic field calculation unit is used to calculate the corresponding compensation electromagnetic field based on the reverse compensation force. A fine-tuning quantity generation unit is used to generate a corresponding control signal fine-tuning quantity based on the compensation electromagnetic field. The control signal output unit is used to superimpose the basic control signal and the control signal fine-tuning amount to obtain the fine-tuned control signal.
4. The motor vibration suppression system based on active magnetic field compensation according to claim 3, characterized in that: It also includes a magnetic field sensor for detecting actual compensated electromagnetic fields; The control device further includes: a magnetic field correction unit; The magnetic field correction unit, after obtaining the control signal fine-tuning amount from the fine-tuning amount generation unit, compares the difference between the actual compensation electromagnetic field and the target compensation electromagnetic field obtained by the compensation electromagnetic field calculation unit, and corrects the control signal fine-tuning amount.
5. The motor vibration suppression system based on active magnetic field compensation according to claim 4, characterized in that: The control device also includes a preprocessing unit; After acquiring the current vibration data, the preprocessing unit performs normalization and filtering noise reduction on the current vibration data, and then outputs the preprocessed current vibration data to the machine learning model and automatic controller.
6. The motor vibration suppression system based on active magnetic field compensation according to any one of claims 1-5, characterized in that: The machine learning model is an LSTM model.
7. A control method applied to the motor vibration suppression system of claim 1, characterized in that, Including the following steps: S1, Generate the initial basic control signal based on historical vibration data; S2, generate a compensation electromagnetic field based on the initial basic control signal, and simultaneously acquire the current vibration data; S3, update the basic control signal based on historical vibration data and current vibration data; S4. Based on the current vibration data, the basic control signal is fine-tuned to obtain the control signal.
8. A control device applied to the motor vibration suppression system of claim 1, characterized in that, include: Machine learning models, feedback units, and automatic controllers; The machine learning model generates initial basic control signals based on historical vibration data; The feedback unit controls the compensation module to generate a compensation electromagnetic field according to the initial basic control signal, and at the same time acquires the current vibration data through the vibration sensor. The machine learning model updates the basic control signal based on historical vibration data and current vibration data; The automatic controller fine-tunes the basic control signal based on the current vibration data to obtain a control signal.