Position sensorless PMSM control method based on improved particle filter algorithm
By improving the particle filtering algorithm and traceless Kalman filtering algorithm, the nonlinear state space model of PMSM is established, which solves the problem of obtaining rotor position and speed under positionless sensor in PMSM, and realizes efficient positionless sensor control.
Patent Information
- Application Number
- CN202510234708.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-28
- Publication Date
- 2025-05-13
AI Technical Summary
In the prior art, without a position sensor in PMSM, it is difficult to accurately obtain rotor position and speed information, especially at the moment of speed increase or initial start-up.
The improved particle filtering algorithm is adopted to establish a nonlinear state space model of PMSM, and the trackless Kalman filtering algorithm is used to estimate the rotor's rotation speed and angle, and closed-loop control is carried out to achieve control without position sensors.
PMSM position sensorless control is realized, which improves the dynamic performance and control accuracy of the system and reduces system losses.
Smart Images
Figure CN119995441A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the technical field of permanent magnet synchronous motor control, and more specifically, to a position sensorless PMSM control method based on an improved particle filter algorithm. Background Art
[0002] Permanent Magnet Synchronous Motor (PMSM), as a new type of energy-saving motor, is gradually replacing traditional asynchronous motors with its high efficiency, high power factor, low loss, low temperature rise, and low noise. For permanent magnet synchronous motors, accurate acquisition of rotor position information is the key to accurate operation and precise control of the motor. In traditional control strategies, position sensors are used to sample speed information and rotor position as feedback for the speed loop and current loop to control the stable operation of the motor. However, position sensors are difficult to install, and once an error occurs, the rotor position information cannot be accurately obtained; installing position sensors will increase the size of the motor and increase system costs; the encoder signal is easily distorted in harsh environments, resulting in reduced accuracy and other problems. In order to increase system reliability and control accuracy and solve the drawbacks of mechanical sensors in obtaining rotor information, the following methods are often used: based on the salient polarity of the motor itself, that is, by injecting high-frequency signals into the motor windings to extract the rotor position contained in the high-frequency current signal; based on the fundamental wave signal, that is, by extracting the rotor signal contained in the fundamental wave. The above two methods are based on the PMSM mathematical model and analyze relevant electrical signals such as voltage and current to extract rotor information, which can get rid of the dependence on mechanical sensors.
[0003] However, the high-frequency signal injection method is mainly used for PMSM zero-low speed sensorless control and has strong robustness; as the speed increases, the system loss is relatively large due to the limitation of the injection signal frequency, which is not conducive to the improvement of the system dynamic performance. When the fundamental signal extraction method extracts position information, the motor speed is very low at the initial startup moment, and the induced electromotive force signal is not obvious, making it difficult to obtain accurate rotor position and speed information. Summary of the invention
[0004] In view of the defects existing in the related technology, the embodiment of the present application provides a position sensorless PMSM control method based on an improved particle filter algorithm, aiming to solve the problem of obtaining the rotor position and speed of the PMSM without a position sensor.
[0005] In a first aspect, an embodiment of the present application provides a position sensorless PMSM control method based on an improved particle filter algorithm, comprising: Obtain the two-phase voltage and two-phase current of the PMSM stator in the two-phase stationary coordinate system; A nonlinear state space model of PMSM is established, with the stator two-phase current, rotor speed and rotor angle as the state vector of the motor system, the stator two-phase voltage as the input vector of the motor system, and the stator two-phase current as the observation vector of the motor system. The nonlinear state space model includes state equations and observation equations. The unscented Kalman filter algorithm is used to estimate the rotor speed and angle; The rotor speed and angle are used as feedback of the control loop to achieve position sensorless control of PMSM.
[0006] In a second aspect, the embodiment of the present application further provides a position sensorless PMSM control device based on an improved particle filter algorithm, characterized in that it includes: An acquisition module, used for acquiring two-phase voltage and two-phase current of the PMSM stator in a two-phase stationary coordinate system; A modeling module is used to establish a nonlinear state space model of the PMSM, taking the stator two-phase current, rotor speed and rotor angle as the state vector of the motor system, taking the stator two-phase voltage as the input vector of the motor system, and taking the stator two-phase current as the observation vector of the motor system. The nonlinear state space model includes a state equation and an observation equation; An estimation module, used for estimating the speed and angle of the rotor using an unscented Kalman filter algorithm; The control module is used to realize position sensorless control of the PMSM by using the speed and angle of the rotor as feedback of the control loop.
[0007] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: at least one memory for storing programs; and at least one processor for executing the programs stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method described in the first aspect or any possible implementation of the first aspect.
[0008] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.
[0009] In a fifth aspect, an embodiment of the present application further provides a computer program product, which, when executed on a processor, enables the processor to execute the method described in the first aspect or any possible implementation of the first aspect.
[0010] The present application provides a position sensorless PMSM control method based on an improved particle filter algorithm, establishes a nonlinear state space model of the PMSM, uses the voltage and current signals detected by the voltage and current sensors as model inputs, and uses the unscented Kalman filter algorithm to estimate the speed and angle of the motor rotor and perform closed-loop control, thereby realizing position sensorless control of the PMSM. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] In order to more clearly illustrate the technical solutions in the present application or related technologies, the following is a brief introduction to the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0012] Figure 1 It is a flow chart of a position sensorless PMSM control method based on an improved particle filter algorithm provided in an embodiment of the present application; Figure 2 It is a schematic diagram of the principle of a position sensorless PMSM control system based on an improved particle filter algorithm provided in an embodiment of the present application; Figure 3 It is a structural schematic diagram of a position sensorless PMSM control device based on an improved particle filter algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION
[0013] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0014] The traditional particle filter algorithm generally has the following situation: after several iterations, except for a few particles, the weight values of most particles will be so small that they can be almost ignored. This can be compensated by increasing the number of particles, resampling technology and choosing a reasonable recommended density distribution function.
[0015] As an improved particle filter algorithm, the Unscented Kalman Filter (UKF) algorithm is suitable for state estimation of nonlinear systems. In the embodiment of the present application, the Unscented Kalman Filter is used to realize PMSM position sensorless control, mainly by finding the optimal suggested density distribution function to make up for the defects of the traditional particle filter algorithm.
[0016] Figure 1 is a flow chart of a position sensorless PMSM control method based on an improved particle filter algorithm provided in an embodiment of the present application, such as Figure 1 As shown, the method comprises at least the following steps: S101. Obtain two-phase voltages and two-phase currents of a PMSM stator in a two-phase stationary coordinate system.
[0017] Specifically, the voltage and current sensors are used to detect the voltage and current signals of the permanent magnet synchronous machine PMSM stator in real time, and Perform coordinate transformation in the two-phase stationary coordinate system to obtain the two-phase voltage of the permanent magnet synchronous machine PMSM stator: and , and the two-phase current and .
[0018] S102. Establish a nonlinear state space model of PMSM, take the stator two-phase current, rotor speed and rotor angle as the state vector of the motor system, take the stator two-phase voltage as the input vector of the motor system, and take the stator two-phase current as the observation vector of the motor system.
[0019] Specifically, in In the two-phase stationary coordinate system, the stator voltage equation of the permanent magnet synchronous machine PMSM satisfies: ; in, and Represent the two-phase voltage of the stator, and Represent the two-phase current of the stator, is the stator resistance, represents the synchronous reactance, and represents the intermediate vector, and denote the rotor speed and angle respectively, Represents the rotor excitation flux. The intermediate vector and satisfy: ; Based on this, the state parameters are selected to establish the nonlinear state space model of the nonlinear permanent magnet synchronous machine PMSM as follows: ① Take the stator two-phase current, rotor speed and rotor angle as the state vector of the motor system: ; ②Use the stator two-phase voltage as the input vector of the motor system: ; ③ Take the stator two-phase current as the observation vector of the motor system: .
[0020] Furthermore, PMSM is established in The nonlinear state space model in a two-phase stationary coordinate system includes the state equation and the observation equation , respectively as follows: ; in, represents the state vector of the motor system at time t, represents the observation vector of the motor system at time t, express The current state value at time t-1 in the two-phase stationary coordinate system, represents the constant coefficient, represents the process noise, represents the constant coefficient, represents the observation noise.
[0021] S103, using an unscented Kalman filter algorithm to estimate the rotation speed and rotation angle of the rotor.
[0022] Specifically, the specific implementation process of estimating the rotor speed and angle using the unscented Kalman filter algorithm is as follows: S1031. Select a particle set at an initial time, and initialize the state estimation and covariance matrix of the particles.
[0023] First, N particles at the initial moment are randomly sampled based on the posterior probability distribution to form a particle set: ; That is, the posterior probability distribution is represented by a particle set.
[0024] Then, initialize the particle state estimate and covariance matrix, including: Initialize the Sigma point set mean of the i-th particle. The Sigma point set mean is an approximate representation of the particle state estimate, reflecting the center position of the state distribution. At the initial moment, the Sigma point set mean of the i-th particle specifically satisfies: ; Initialize the covariance matrix of the i-th particle to satisfy: The particle set satisfies the posterior probability distribution, so the posterior probability distribution value of the i-th particle at the initial moment, that is, the covariance matrix, is: ; in, Represents the number of Sigma points in the Sigma point set, and Represents process noise and observation noise The variance of , and the mean is 0. The value of is generally , Represents the state vector dimension.
[0025] S1032. Generate a Sigma point set of each particle at time k-1 through untraceable transformation according to the state estimation and covariance matrix of each particle at time k-1.
[0026] The Sigma point set of the particle is symmetrically distributed around the current state estimate and covariance matrix. dimensional state vector, usually generated Sigma points, the Sigma point set of the i-th particle at time k-1 specifically satisfies: ; in, represents the state vector dimension, Represents the scaling factor used to reduce the overall prediction error.
[0027] S1033. Propagate Sigma points through the state equation to perform state prediction.
[0028] The specific process of state prediction is: For i of particles One-step prediction is performed on Sigma points, satisfying: ; Calculate the covariance matrix of the one-step forecast, satisfying: ; in, Represents the weight of the j-th Sigma point set.
[0029] S1034. Propagate Sigma points through observation equations to make observation predictions.
[0030] The specific process of observation and prediction is as follows: Substitute the one-step prediction value into the observation equation to obtain the observed value that satisfies: ; Perform a weighted sum of the observations, and the weighted sum value 1 satisfies: ; The predicted mean and predicted covariance matrix of the observed prediction are determined based on the weighted sum value. The predicted mean satisfies: ; The prediction covariance matrix satisfies: .
[0031] S1035. Calculate the Kalman gain based on the mean and covariance matrix of the observation prediction, and update the state estimation and covariance matrix of the particle at time k.
[0032] The Kalman gain matrix is: ; The particle state is estimated to be: ; The particle covariance matrix is: .
[0033] Update the state of the particle according to the new observation value to satisfy: ; in, represents a Gaussian distribution, , .
[0034] Furthermore, normalize the particles and recalculate the weights. The updated particle weights are: .
[0035] S104, using the rotation speed and rotation angle of the rotor as feedback of the control loop to achieve position sensorless control of the PMSM.
[0036] Specifically, the UKF-based state estimator estimates the rotation speed and rotation angle of the rotor, that is, the rotation speed and position of the rotor in real time. Figure 2 is a schematic diagram of the principle of a position sensorless PMSM control system based on an improved particle filter algorithm provided in an embodiment of the present application, such as Figure 2 As shown, current loop control is implemented according to the estimated rotor position and speed; and speed loop control is implemented by adjusting the motor speed according to the estimated rotor speed.
[0037] The embodiment of the present application provides a position sensorless PMSM control method based on an improved particle filter algorithm, establishes a nonlinear state space model of the PMSM, uses the voltage and current signals detected by the voltage and current sensors as model inputs, and uses the unscented Kalman filter algorithm to estimate the speed and angle of the motor rotor and perform closed-loop control, thereby realizing position sensorless control of the PMSM.
[0038] Figure 3 is a structural diagram of a position sensorless PMSM control device based on an improved particle filter algorithm provided in an embodiment of the present application, such as Figure 3 As shown, the device at least includes: An acquisition module 301 is used to acquire two-phase voltage and two-phase current of a PMSM stator in a two-phase stationary coordinate system; A modeling module 302 is used to establish a nonlinear state space model of the PMSM, using the stator two-phase current, the rotor speed and the rotor angle as the state vector of the motor system, the stator two-phase voltage as the input vector of the motor system, and the stator two-phase current as the observation vector of the motor system, wherein the nonlinear state space model includes a state equation and an observation equation; An estimation module 303, for estimating the speed and angle of the rotor using an unscented Kalman filter algorithm; The control module 304 is used to use the rotation speed and rotation angle of the rotor as feedback of the control loop to achieve position sensorless control of the PMSM.
[0039] In some embodiments, the estimation module 303 is specifically configured to: Select the particle set at the initial moment and initialize the particle state estimation and covariance matrix; According to the state estimation and covariance matrix of the particle at time k-1, the Sigma point set of the particle at time k-1 is generated by unscented transformation; Propagate Sigma points through the state equation to predict the state; Propagate Sigma points through observation equations to make observation predictions; The Kalman gain is calculated based on the mean and covariance matrix of the observation prediction, and the state estimate and covariance matrix of the particle at time k are updated.
[0040] In some embodiments, selecting a particle set at an initial time includes: Based on the posterior probability distribution, N particles are randomly sampled at the initial moment to form a particle set: ; Initialize the particle state estimate and covariance matrix, including: Initialize the mean of the particle's Sigma point set to satisfy: ; Initialize the particle covariance matrix to satisfy: ; in, i represents the particle number, Represents the number of Sigma points in the Sigma point set, and Represents process noise and observation noise The variance of , and the mean is 0.
[0041] In some embodiments, the Sigma point set of the i-th particle at time k-1 satisfies: ; in, represents the state vector dimension, Indicates the zoom ratio.
[0042] In some embodiments, the state prediction is performed by propagating Sigma points through the state equation, including: right One-step prediction is performed on Sigma points, satisfying: ; Calculate the covariance matrix of the one-step forecast, satisfying: ; in, Represents the weight of the j-th Sigma point.
[0043] In some embodiments, propagating Sigma points through observation equations to perform observation predictions includes: Substitute the one-step prediction value into the observation equation to obtain the observed value, which satisfies: ; Perform a weighted summation on the observations to satisfy: ; The predicted mean and predicted covariance matrix of the observed prediction are determined based on the weighted sum of the observed values. The predicted mean satisfies: , the prediction covariance matrix satisfies: .
[0044] In some embodiments, the Kalman gain satisfies: .
[0045] In some embodiments, the state estimate of the ith particle after updating satisfies: ; After updating, the covariance matrix of the i-th particle satisfies: ; After the update, the state of the i-th particle satisfies: .
[0046] In some embodiments, the nonlinear state space model of the PMSM satisfies: ; in, represents the state vector of the motor system at time t, represents the observation vector of the motor system at time t, express The current state value at time t-1 in the two-phase stationary coordinate system, is the stator resistance, represents the synchronous reactance, represents the rotor excitation flux, and denote the rotor speed and angle respectively, and Represent the two-phase current of the stator, represents the constant coefficient, Indicates the voltage state parameter, represents the process noise, and Represent the two-phase voltage of the stator, represents the constant coefficient, represents the observation noise.
[0047] It can be understood that the detailed functional implementation of each of the above-mentioned units / modules can be found in the introduction of the aforementioned method embodiment, and will not be repeated here.
[0048] It should be understood that the above-mentioned device is used to execute the method in the above-mentioned embodiment. The implementation principle and technical effect of the corresponding program module in the device are similar to those described in the above-mentioned method. The working process of the device can refer to the corresponding process in the above-mentioned method, which will not be repeated here.
[0049] Based on the method in the above embodiment, an embodiment of the present application provides an electronic device. The device may include: at least one memory for storing programs and at least one processor for executing the programs stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method described in the above embodiment.
[0050] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method in the above embodiment.
[0051] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method in the above embodiment.
[0052] It is understandable that the processor in the embodiment of the present application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.
[0053] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, and the software modules can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC.
[0054] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.
[0055] It should be understood that the various numerical numbers involved in the embodiments of the present application are only used for the convenience of description and are not used to limit the scope of the embodiments of the present application.
[0056] It is easy for those skilled in the art to understand that the above are only preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of protection of the present application.
Claims
1. A position sensorless PMSM control method based on an improved particle filter algorithm, characterized in that: include: Obtain the two-phase voltage and two-phase current of the PMSM stator in the two-phase stationary coordinate system; Establishing a nonlinear state space model of PMSM, taking the stator two-phase current, rotor speed and rotor angle as the state vector of the motor system, taking the stator two-phase voltage as the input vector of the motor system, and taking the stator two-phase current as the observation vector of the motor system, wherein the nonlinear state space model includes a state equation and an observation equation; The unscented Kalman filter algorithm is used to estimate the rotor speed and angle; The rotor speed and angle are used as feedback of the control loop to achieve position sensorless control of PMSM.
2. The position sensorless PMSM control method according to claim 1, characterized in that: The method of estimating the rotation speed and rotation angle of the rotor by using the unscented Kalman filter algorithm includes: Select the particle set at the initial moment and initialize the particle state estimation and covariance matrix; According to the state estimation and covariance matrix of the particle at time k-1, the Sigma point set of the particle at time k-1 is generated by unscented transformation; Propagate Sigma points through the state equation to perform state prediction; Propagate Sigma points through the observation equation to make observation predictions; The Kalman gain is calculated based on the mean and covariance matrix of the observation prediction, and the state estimate and covariance matrix of the particle at time k are updated.
3. The position sensorless PMSM control method according to claim 2, characterized in that: The particle set selected at the initial moment includes: Based on the posterior probability distribution, N particles are randomly sampled at the initial moment to form a particle set: ; The initialization of the particle state estimation and covariance matrix includes: Initialize the mean of the particle's Sigma point set to satisfy: ; Initialize the particle covariance matrix to satisfy: ； in, i represents the particle number, Represents the number of Sigma points in the Sigma point set, and Represents process noise and observation noise The variance of , and the mean is 0.
4. The position sensorless PMSM control method according to claim 3, characterized in that: The Sigma point set of the particle at time k-1 satisfies: ； in, represents the state vector dimension, Indicates the zoom ratio.
5. The position sensorless PMSM control method according to claim 4, characterized in that: The state prediction is performed by propagating Sigma points through the state equation, including: right One-step prediction is performed on Sigma points, satisfying: ； Calculate the covariance matrix of the one-step forecast, satisfying: ； in, Represents the weight of the j-th Sigma point.
6. The position sensorless PMSM control method according to claim 5, characterized in that: The propagating Sigma points through the observation equation to perform observation prediction includes: Substitute the one-step prediction value into the observation equation to obtain the observation value, which satisfies: ; Perform a weighted summation on the observations to satisfy: ; The predicted mean and predicted covariance matrix of the observed prediction are determined based on the weighted sum of the observed values, and the predicted mean satisfies: , the prediction covariance matrix satisfies: .
7. The position sensorless PMSM control method according to claim 6, characterized in that: The Kalman gain satisfies: .
8. The position sensorless PMSM control method according to claim 7, characterized in that: After the update, the state estimate of the i-th particle satisfies: ; After updating, the covariance matrix of the i-th particle satisfies: ; After the update, the state of the i-th particle satisfies: .
9. The position sensorless PMSM control method according to claim 1, characterized in that: The nonlinear state space model of the PMSM satisfies: ； in, represents the state vector of the motor system at time t, represents the observation vector of the motor system at time t, express The current state value at time t-1 in the two-phase stationary coordinate system, is the stator resistance, represents the synchronous reactance, represents the rotor excitation flux, and denote the rotor speed and angle respectively, and Represent the two-phase current of the stator, represents the constant coefficient, Indicates the voltage state parameter, represents the process noise, and Represent the two-phase voltage of the stator, represents the constant coefficient, represents the observation noise.
10. A position sensorless PMSM control device based on an improved particle filter algorithm, characterized in that: include: An acquisition module, used for acquiring two-phase voltage and two-phase current of the PMSM stator in a two-phase stationary coordinate system; A modeling module, used to establish a nonlinear state space model of the PMSM, using the stator two-phase current, the rotor speed and the rotor angle as the state vector of the motor system, the stator two-phase voltage as the input vector of the motor system, and the stator two-phase current as the observation vector of the motor system, wherein the nonlinear state space model includes a state equation and an observation equation; An estimation module, used for estimating the speed and angle of the rotor using an unscented Kalman filter algorithm; The control module is used to realize position sensorless control of the PMSM by using the speed and angle of the rotor as feedback of the control loop.