A permanent magnet propulsion motor anti-interference control method and system based on inertia identification and disturbance error constraint enhanced self-anti-interference

Abstract

This invention discloses a disturbance rejection control method and system for permanent magnet propulsion motors based on inertia identification and disturbance error constraint-enhanced active disturbance rejection. In the design of the active disturbance rejection controller, an extended state observer is cascaded, and a disturbance error constraint is introduced into the observer's disturbance estimation constraint, forming a disturbance error constraint-enhanced active disturbance rejection controller. This improves the controller's disturbance rejection performance when the controlled object faces parameter perturbations and load disturbances. Compared with traditional active disturbance rejection controllers, the designed disturbance error constraint-enhanced active disturbance rejection controller has significantly enhanced disturbance rejection capability. Employing a model reference adaptive algorithm with self-correcting gain function, the controller gain accuracy is improved by real-time identification of the moment of inertia of the permanent magnet propulsion motor control system and real-time adjustment of the controller gain. This will enhance the control performance of the permanent magnet propulsion motor under unknown operating conditions.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet propulsion motor control, and in particular to a disturbance rejection control method and system for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement. Background Technology

[0002] As the core power source for underwater equipment, the technological level of underwater propulsion motors directly affects the performance and reliability of various underwater vehicles and operational equipment. Permanent magnet synchronous motors (PMSMs) possess advantages such as high torque density, high efficiency, good speed regulation performance, and low noise, making them suitable for high-performance applications and widely used in the propulsion motor bodies of various underwater equipment.

[0003] In the vector control system of permanent magnet synchronous motor, the speed loop and the current loop form a double closed-loop structure. The speed loop plays a key role in the system's operating performance and is usually regulated by a PI controller. However, using a traditional PI controller for the speed loop can lead to problems such as slow dynamic response and significant speed overshoot during speed regulation. In addition, traditional PI control struggles to balance system stability and anti-interference capabilities, requiring a trade-off between parameter selection, and is ill-suited to handle complex underwater flow disturbances.

[0004] Compared to traditional PI control, active disturbance rejection (ADRF) offers advantages such as low dependence on the system model, rapid response without overshoot, and strong disturbance rejection performance. In underwater environments, ADRF treats internal and external disturbances, including fluid response and parameter changes, as a single total disturbance, utilizing an extended state observer for observational compensation to achieve strong disturbance rejection in the drive control system and meet the high-performance control requirements of the vessel. For the extended state observer, a larger bandwidth results in smaller observation errors and better system disturbance rejection. However, due to high-frequency noise in the sensors, the bandwidth of the extended state observer needs to be limited. This bandwidth limitation restricts the disturbance rejection capability of the ADRF controller. To improve the disturbance rejection capability of the ADRF controller, improvements to traditional ADRF methods are necessary.

[0005] For example, the improved method disclosed in Chinese patent application No. 202410971251.8, entitled "Design Method of Cascaded Active Disturbance Rejection Controller for Ship Propulsion Permanent Magnet Motor," designs a three-stage cascaded active disturbance rejection controller by cascading a traditional linear extended state observer and a traditional nonlinear extended state observer, and introducing a nonlinear error feedback control law. This improves the accuracy of the observer's estimation of load disturbances and parameter perturbations, as well as the response speed and disturbance rejection capability of the active disturbance rejection controller. However, this method uses a nonlinear extended state observer and a nonlinear error feedback control law, resulting in a complex controller structure and difficult parameter tuning. In underwater equipment, propulsion motors are usually directly or indirectly connected to propellers, making it difficult to obtain accurate measurement and calculation results for the "control gain" of the controlled system, a crucial parameter of the active disturbance rejection controller in the motor control system. Therefore, the mismatch in control gain leads to a decrease in the control performance of the propulsion motor control system. Specifically, for the speed loop active disturbance rejection controller, the control gain is related to the torque coefficient and the moment of inertia. Meanwhile, disturbances in complex environments such as underwater pressure, temperature difference, and unknown water flow can also cause changes in the motor's moment of inertia, making it difficult to accurately match the system's control gain and affecting the system's control performance.

[0006] To address these issues, the improved linear active disturbance rejection controller of this invention has advantages such as simple parameter tuning and strong anti-interference capability. Combined with rotational inertia identification technology, the rotational inertia is identified and the identification result is fed back to the active disturbance rejection controller to adjust the system control gain value in real time, ultimately achieving strong anti-interference capability of the propulsion motor control system. Summary of the Invention

[0007] The purpose of this invention is to provide a disturbance rejection control method and system for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement. During use, this method can solve the problem of large control gain error caused by inertia mismatch and improve the anti-interference capability of the speed regulation system.

[0008] The technical solution to achieve the purpose of this invention is: a disturbance rejection control method for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement, comprising the following steps:

[0009] Step 1: In the permanent magnet motor control system, the rotor mechanical angle θ of the permanent magnet motor 1 is acquired by using encoder module 2 and current sampling module 4. m and three-phase current i a i b i c ;wherein the rotor mechanical angle θ m Then, the rotor angular velocity ω is obtained through differential module 3. m and rotor mechanical angle θ m and the number of pole pairs p of the motorn Multiplication yields the rotor electrical angle θ e The three-phase current i a i b i c The current i in the two-phase stationary coordinate system is obtained through Clark coordinate transformation module 5. α i β Then, the current i in the two-phase stationary coordinate system α i β and rotor electrical angle θ e The signal is input into Park coordinate transformation module 6 to obtain the quadrature and direct axis current signal i. q i d ;

[0010] Step 2: The quadrature-axis current i obtained by the Park coordinate transformation module 6 is... q With torque coefficient K t Multiplication yields the electromagnetic torque T of the motor. e Then, the rotor angular velocity ω obtained by the differential module 3 is... m and the electromagnetic torque T of the motor e The estimated value of the motor's rotational inertia is obtained by inputting the model reference inertia identification algorithm module 7, which is a self-calibrating gain. ;

[0011] Step 3: The estimated value of the motor rotational inertia obtained by the self-calibrating gain model reference inertia identification algorithm module 7 is used. The desired rotor angular velocity ω output by module 8 is the desired rotational speed setting. ref The rotor angular velocity ω output by differential module 3 m The input is fed into the disturbance error constraint-enhanced active disturbance rejection controller module 9, and after calculation by the disturbance error constraint-enhanced active disturbance rejection controller module 9, the initial desired q-axis current i is output. q * The desired q-axis current i is obtained through the limiting module 10. qsat * ;

[0012] Step 4: The desired q-axis current i output by the limiting module 10... qsat * The actual q-axis current i output by Park coordinate transformation module 6 q By subtraction, the q-axis current deviation e is obtained. iq , the q-axis current deviation e iq The input is fed into PI module 12 to obtain the desired q-axis voltage u. q Similarly, the i output by module 11, which is the desired current setting module for the d-axis, is... d * The actual d-axis current i output by Park coordinate transformation module 6d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into PI module 13 to obtain the desired d-axis voltage u. d ;

[0013] Step 5: Set the desired d-axis voltage u d q-axis desired voltage u q and rotor electrical angle θ e The input is fed into the inverse Park coordinate transformation module 14 to obtain the desired voltage u in the two-phase stationary coordinate system. α u β and u α u β The switching signal S, used to drive the inverter module 16, is obtained as input to the space vector pulse width modulation module 15. A S B S C Inverter module 16 receives switching signal S A S B S C Then, the drive signal of the motor is obtained, which enables the permanent magnet motor 1 to work, forming a closed-loop control.

[0014] The beneficial effects of this invention are:

[0015] 1. In the design of an active disturbance rejection controller (ADRC), this invention employs a cascaded design for the extended state observer and introduces a disturbance error constraint into the observer disturbance estimation constraint, forming a disturbance error constraint-enhanced ADRC. This improves the controller's disturbance rejection performance when the controlled object faces parameter perturbations and load disturbances. Compared with traditional ADRCs, the designed disturbance error constraint-enhanced ADRC exhibits significantly enhanced disturbance rejection capability.

[0016] 2. This invention employs a model reference adaptive algorithm with self-correcting gain function. By identifying the rotational inertia of the permanent magnet propulsion motor control system in real time and adjusting the controller gain in real time, it achieves the effect of improving the gain accuracy of the control system. This will improve the control performance of the permanent magnet propulsion motor when operating under unknown conditions. Attached Figure Description

[0017] Figure 1 This is a structural block diagram of the anti-disturbance control system for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement according to the present invention.

[0018] Figure 2 for Figure 1 The structural block diagram of module 7, which is the model reference inertia identification algorithm module for self-calibrating gain.

[0019] Figure 3 for Figure 1 Block diagram of the structure of the enhanced active disturbance rejection controller module 9 with disturbance error constraint.

[0020] Figure 4 The diagram shows a comparison of the rotational speed waveforms of propulsion systems based on a PI controller, a conventional linear active disturbance rejection controller (LADRC), a disturbance error constrained active disturbance rejection controller (DADRC), and a disturbance error constrained enhanced active disturbance rejection controller (DSADRC) in embodiments of the present invention.

[0021] Figure 5 The images show the identification results of the traditional model reference inertia identification algorithm with q=0.002, the traditional model reference inertia identification algorithm with q=0.022, and the model reference adaptive inertia identification algorithm with self-correcting gain in the embodiments of the present invention.

[0022] Figure 6 This is a speed tracking diagram of the permanent magnet propulsion motor anti-disturbance control system based on inertia identification and disturbance error constraint enhancement in an embodiment of the present invention when the speed command is a square wave.

[0023] Figure 7 The diagram shows the rotational inertia identification results of the permanent magnet propulsion motor anti-disturbance control system based on inertia identification and disturbance constraint enhancement in this embodiment of the invention when the rotational speed is given as a square wave.

[0024] In the diagram: 1. Permanent magnet motor; 2. Encoder module; 3. Differentiator module; 4. Current sampling module; 5. Clark coordinate transformation module; 6. Park coordinate transformation module; 7. Model reference inertia identification algorithm module with self-calibrating gain; 8. Desired speed setting module; 9. Disturbance error constraint enhanced active disturbance rejection controller module; 10. Limiting module; 11. d-axis desired current setting module; 12. PI module; 13. PI module; 14. Inverse Park coordinate transformation module; 15. Space vector pulse width modulation module. ; 16. Inverter module; 17. Discretization module; 18. Reference model module; 19. Adjustable model module; 20. Rotational inertia identification sub-model module; 21. Delayed sampling module; 22. Self-calibration mechanism module; 23. Delayed sampling module; 24. Continuous processing module; 25. First-level disturbance error constraint extended state observer module; 26. Second-level disturbance error constraint extended state observer module; 27. Proportional control law module; 28. Control gain calculation module; 29. ​​Initial desired current calculation module. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0026] like Figure 1 As shown, this invention relates to a disturbance rejection control system for a permanent magnet propulsion motor based on inertia identification and disturbance error constraint-enhanced active disturbance rejection. This control system comprises a permanent magnet motor 1, an encoder module 2, a differential module 3, a current sampling module 4, a Clark coordinate transformation module 5, a Park coordinate transformation module 6, a model reference inertia identification algorithm module with self-correcting gain 7, a desired speed setting module 8, a disturbance error constraint-enhanced active disturbance rejection controller module 9, a limiting module 10, a d-axis desired current setting module 11, a PI module 12, a PI module 13, an inverse Park coordinate transformation module 14, a space vector pulse width modulation module 15, and an inverter module 16. PI modules 12 and 13 both employ PI controllers.

[0027] During the disturbance rejection control system operation, the rotor mechanical angle θ of the permanent magnet motor 1 is acquired by using encoder module 2 and current sampling module 4. m and three-phase current i a i b i c ;wherein the rotor mechanical angle θ m The rotor angular velocity ω is obtained through differential module 3. m and rotor mechanical angle θ m and the number of pole pairs p of the motor n Multiplication yields the rotor electrical angle θ e The three-phase current i a i b i c The current i in the two-phase stationary coordinate system is obtained through Clark coordinate transformation module 5. α i β Then, the current i in the two-phase stationary coordinate system α i β and rotor electrical angle θ e The signal is input into Park coordinate transformation module 6 to obtain the quadrature and direct axis current signal i. q i d The quadrature-axis current i obtained through Park coordinate transformation module 6 q With torque coefficient K t Multiplication yields the electromagnetic torque T of the motor. e Then, the rotor angular velocity ω obtained by the differential module 3 is... m and the electromagnetic torque T of the motor e The estimated value of the motor's rotational inertia is obtained by inputting the model reference inertia identification algorithm module 7, which is a self-calibrating gain. The estimated value of the motor's rotational inertia obtained by the model reference inertia identification algorithm module 7 after correction of gain is used. The desired rotor angular velocity ω output by module 8 is the desired rotational speed setting.ref The mechanical angular velocity ω output by differential module 3 m The input is fed into the disturbance error constraint-enhanced active disturbance rejection controller module 9, and after calculation by the disturbance error constraint-enhanced active disturbance rejection controller module 9, the initial desired q-axis current i is output. q * After being limited by the limiting module 10 (which limits the output value to a set range), the desired q-axis current i is obtained. qsat * The desired q-axis current i output by the limiting module 10 is... qsat * The actual q-axis current i output by Park coordinate transformation module 6 q By subtraction, the q-axis current deviation e is obtained. iq , the q-axis current deviation e iq The input is fed into PI module 12 to obtain the desired q-axis voltage u. q Similarly, the i output by the d-axis desired current module 11 will be... d * The actual d-axis current i output by Park coordinate transformation module 6 d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into PI module 13 to obtain the desired d-axis voltage u. d ; the desired voltage u along the d-axis d q-axis desired voltage u q and rotor electrical angle θ e The input is fed into the inverse Park coordinate transformation module 14 to obtain the desired voltage u in the two-phase stationary coordinate system. α u β The desired voltage u obtained α u β The input is sent to the space vector pulse width modulation module 15 to obtain the switching signal S used to drive the inverter module 16. A S B S C The switching signal is input to the three-phase inverter module 16 to generate a pulse signal, thereby realizing the drive control of the permanent magnet motor 1.

[0028] First, the motor parameters are obtained from the motor nameplate. The parameters of the permanent magnet motor used in this embodiment of the invention are: rated speed of 500 rpm / min, inductance L... d The inductance is 3.245mH, and the inductance L is... q The stator resistance is 3.75mH. s The permanent magnet flux linkage ψ is 0.58Ω. f It is 0.115 Wb, and the pole pair number is p. nThe value is 7, and the moment of inertia J is 0.00118 kg·m^2. During the entire motor control process, the sampling time T... s Set to 1×10 -4 s.

[0029] like Figure 2 As shown, the self-calibrating gain model reference inertia identification algorithm module 7 consists of a discretization processing module 17, a reference model module 18, an adjustable model module 19, a rotational inertia identification sub-model module 20, a delayed sampling module 21, a self-calibration mechanism module 22, a delayed sampling module 23, and a continuousization module 24. Figure 1 Cross-axis current i q With torque coefficient K t The electromagnetic torque T of the motor obtained by multiplication e Japanese Classics Figure 1 The rotor angular mechanical velocity ω obtained by the differential module 3 m enter Figure 2 In the discretization module 17 of the self-calibrating gain model reference inertia identification algorithm module 7, the electromagnetic torque T at time k-1 is obtained through the discretization processing module 17. e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the mechanical angular velocity ω of the motor at time k-1. m The mechanical angular velocity ω of the motor at time (k-1) and time (k-2) m (k-2); The electromagnetic torque T at time k-1 output by the discretization processing module 17. e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the mechanical angular velocity ω of the motor at time k-1. m The mechanical angular velocity ω of the motor at time (k-1) and time (k-2) m (k-2) is input into the reference model module 18, and the mechanical angular velocity ω of the motor at time k is obtained through the reference model module 18. m (k), whose expression is:

[0030]

[0031] In the formula, ω m (k), ω m (k-1), ω m (k-2) represents the mechanical angular velocity of the motor at times k, k-1, and k-2, respectively, and T e (k-1), T e (k-2) represents the electromagnetic torque of the motor at times k-1 and k-2, T s Where is the sampling time, J is the motor's moment of inertia, and h is the identification parameter.

[0032] At the same time, the electromagnetic torque T at the (k-1)th time output by the discretization processing module 17 will be... e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the mechanical angular velocity ω of the motor at time k-1. m (k-1), the mechanical angular velocity ω of the motor at time k-2 m (k-2) and the identification value of the identification parameter at time k-1 output by the delayed sampling module 23. (k-1) is input to the adjustable model module 19, and the estimated mechanical angular velocity of the motor at time k is obtained through the adjustable model module 19. Its expression is

[0033]

[0034] In the formula, This is the estimated rotational speed at time k. (k-1) is the identification value of h at time k-1.

[0035] ω at time k will be obtained through reference model module 18. m (k) The k-th time obtained through the adjustable model module 19 The electromagnetic torque T at time k-1 obtained by discretization module 17 e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the identification value of the identification parameter at time k-1 output by the delayed sampling module 23. (k-1) and the adaptive coefficient q(k) output by the self-calibration mechanism module 22 at time k are used as inputs to the moment of inertia identification sub-model module 20 to calculate the identification value at time k. (k) and the identified value of the moment of inertia at time k Its expression is:

[0036]

[0037] In the formula, q is the adaptive coefficient, and q(k) is the adaptive coefficient at time k. (k) is the identification value of h at time k. Let be the identified value of the moment of inertia at time k.

[0038] Then, the identified moment of inertia at time k is... as well as The rotational inertia identification value at time k-1 obtained by the delayed sampling module 21 The input is fed into the self-calibration mechanism module 22 to obtain the adaptive coefficient q(k) at time k, whose expression is:

[0039]

[0040] In the formula, v is the correction intensity coefficient (in the simulation of this embodiment, the correction intensity coefficient v is set to 10). , e represents the moment of inertia at times k and k-1. J (k) represents the inertia identification velocity at time k, and q0 represents the initial adaptive coefficient (the initial adaptive coefficient q0 is set to 0.002 in the simulation of this embodiment).

[0041] The adaptive coefficient q(k) obtained at time k is input into the rotational inertia identification sub-model module 20 to adjust the rotational inertia identification speed and accuracy of the inertia identification sub-model module 20. The rotational inertia identification value at time k is then used as the input. The input is fed into the continuous processing module 24, and after continuous processing, an estimated value of the moment of inertia is obtained. .

[0042] Furthermore, it will be through Figure 1 The rotor angular velocity ω obtained by the differential module 3 m The desired rotor angular velocity ω is given by module 8 and the desired rotational speed. ref and through Figure 2 The estimated value of the moment of inertia obtained by the continuous processing module 24 In module 9 of the enhanced active disturbance rejection controller with input disturbance error constraint. For example... Figure 3 As shown, the disturbance error constraint enhanced active disturbance rejection controller module 9 is composed of a first-level disturbance error constraint extended state observer module 25, a second-level disturbance error constraint extended state observer module 26, a proportional control law module 27, a control gain calculation module 28, and an initial desired current calculation module 29.

[0043] Then through Figure 1 The mechanical angular velocity ω obtained by the differential module 3 m Japanese Classics Figure 3 The control quantity u2 obtained from the control quantity calculation and compensation stage is input to the first-level disturbance error constraint expansion state observer module 25. The first-level disturbance error constraint expansion state observer module 25 is used to observe the rotor angular velocity observation value Z1 and the disturbance observation value Z2 of the first-level observer. The specific expression of the first-level disturbance error constraint expansion state observer is as follows:

[0044]

[0045] In the formula, Z1 is the observed rotor angular velocity of the first-stage observer, Z2 is the observed disturbance value of the first-stage observer, and E Z1E represents the velocity observation error of the first-stage observer. Z2 l1 represents the perturbation observation error of the first-stage observer, and l2 represents the observer coefficients, which are greater than 0, where l1 = ω. 01 l2=ω 01 ω 01 The bandwidth of the first-stage observer (ω, the bandwidth of the first-stage observer in the simulation of this embodiment) 01 (Set to 800 rad / s).

[0046] Next, the disturbance observation value Z2 observed by the first-level disturbance error constraint expansion state observer module 25 and the rotor angular velocity ω obtained by the differential module 3 are combined. m Together with the control quantity u2 obtained through the control quantity calculation and compensation process, the second-level disturbance error constraint extended state observer module 26 is input into the second-level disturbance error constraint extended state observer module 26. The disturbance observation value S2 of the second-level observer is observed using the second-level disturbance error constraint extended state observer module 26. The specific expression of the second-level disturbance error constraint extended state observer is as follows:

[0047]

[0048] In the formula, S1 is the observed rotor angular velocity of the second-stage observer, S2 is the observed disturbance value of the second-stage observer, and E S1 E represents the velocity observation error of the second-stage observer. S2 l3 represents the perturbation observation error of the second-stage observer, and l4 represents the observer coefficients, which are greater than 0, where l3 = ω 02 l4=ω 02 ω 02 The bandwidth of the second-stage observer (ω, the bandwidth of the second-stage observer in the simulation of this embodiment of the invention) 02 (Set to 800 rad / s).

[0049] Then, Figure 1 The desired rotor angular velocity ω output by module 8 is the desired rotational speed setting module. ref The speed tracking deviation e is obtained by subtracting the mechanical angular velocity observation value Z1 observed by the first-level disturbance error constraint extended state observer module 25. s The obtained speed tracking deviation e s The input is fed into the proportional control law module 27, where the control quantity u1 is calculated. Its expression is:

[0050]

[0051] In the formula, ω ref e represents the desired rotor angular velocity. s k represents the speed tracking deviation. p This represents the proportional feedback coefficient (the proportional feedback coefficient k in the simulation of this embodiment of the invention).p (Set to 250 rad / s), u1 represents the control quantity without disturbance compensation.

[0052] Next, the proportional control law module 27 calculates the control quantity u1 for disturbance compensation, and adds it to the disturbance observation value Z2 of the first-level observer output by the first-level disturbance error constraint extended state observer module 25 and the disturbance observation value S2 of the second-level observer output by the second-level disturbance error constraint extended state observer module 26 to obtain the total observed disturbance. By taking the difference, we can calculate the control quantity u2 after disturbance compensation, and its expression is as follows:

[0053]

[0054] In the formula, u1 represents the total observed disturbance, and u2 represents the control quantity after disturbance compensation of u1.

[0055] The disturbance-compensated control quantity u2 is fed into the first-stage disturbance error constraint extended state observer module 25 and the second-stage disturbance error constraint extended state observer module 26, and combined with the rotor angular velocity ω input to the two observers. m The rotor angular velocity Z1 and disturbance Z2 of the first-stage observer and the disturbance S2 of the second-stage observer are observed using the observer. Simultaneously, the disturbance-compensated control quantity u2 and the control gain b1 output by the control gain calculation module 28 are used as inputs to the initial desired current calculation module 29, which calculates the initial desired current i. q * .

[0056] Will pass Figure 2 The estimated value of the moment of inertia obtained by the continuous processing module 24 As Figure 3 The control gain b1 of the control system can be calculated by inputting into the control gain calculation module 28, and its expression is:

[0057]

[0058] In the formula, K t p represents the torque coefficient of a permanent magnet motor. n ψ represents the number of pole pairs of the motor. f b1 represents the permanent magnet flux linkage of the motor, and b1 represents the calculated control gain of the control system.

[0059] The calculated control gain b1 of the control system and the control quantity u2 output by the control quantity calculation and compensation loop are jointly input to the initial expected current calculation module 29, and the initial expected current i is obtained using the initial expected current calculation module 29.q * And use it as the output of the disturbance error constraint-enhanced active disturbance rejection controller module 9, with the initial desired current i q * The expression is:

[0060]

[0061] In the formula, i q * This represents the initial expected current.

[0062] Furthermore, the initial desired current i output by the disturbance error constraint-enhanced active disturbance rejection controller module 9 is... q * The desired q-axis current i output after reaching the limiting module 10 qsat * The desired q-axis current i output by the limiting module 10 qsat * The actual q-axis current i output by Park coordinate transformation module 6 q By subtraction, the q-axis current deviation e is obtained. iq , the q-axis current deviation e iq The input is fed into PI module 12 to obtain the desired q-axis voltage u. q Similarly, the i output by the d-axis desired current module 11 will be... d * The actual d-axis current i output by Park coordinate transformation module 6 d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into PI module 13 to obtain the desired d-axis voltage u. d ; the desired voltage u along the d-axis d q-axis desired voltage u q The input is fed into the inverse Park coordinate transformation module 14 to obtain the desired voltage u in the two-phase stationary coordinate system. α u β and u α u β The switching signal S, used to drive the inverter module 16, is obtained as input to the space vector pulse width modulation module 15. A S B S C Inverter module 16 receives switching signal S A S B S C Then, the drive signal of the motor is obtained, thereby enabling the permanent magnet motor 1 to work.

[0063] The disturbance rejection control system for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement comprises a permanent magnet motor 1, an encoder module 2, a differential module 3, a current sampling module 4, a Clark coordinate transformation module 5, a Park coordinate transformation module 6, a model reference inertia identification algorithm module with self-correcting gain 7, a desired speed setting module 8, a disturbance error constraint enhancement-type active disturbance rejection controller module 9, a limiting module 10, a d-axis desired current setting module 11, a PI module 12, a PI module 13, an inverse Park coordinate transformation module 14, a space vector pulse width modulation module 15, and an inverter module 16. During operation, the disturbance rejection control system uses the encoder module 2 and the current sampling module 4 to acquire the rotor mechanical angle θ of the permanent magnet motor 1. m and three-phase current i a i b i c ;wherein the rotor mechanical angle θ m The rotor angular velocity ω is obtained through differential module 3. m and rotor mechanical angle θ m and the number of pole pairs p of the motor n Multiplication yields the rotor electrical angle θ e The three-phase current i a i b i c The current i in the two-phase stationary coordinate system is obtained through Clark coordinate transformation module 5. α i β Then, the current i in the two-phase stationary coordinate system α i β and rotor electrical angle θ e The signal is input into Park coordinate transformation module 6 to obtain the quadrature and direct axis current signal i. q i d The quadrature-axis current i obtained through Park coordinate transformation module 6 q With torque coefficient K t Multiplication yields the electromagnetic torque T of the motor. e Then, the rotor angular velocity ω obtained by the differential module 3 is... m and the electromagnetic torque T of the motor e The estimated value of the motor's rotational inertia is obtained by inputting the model reference inertia identification algorithm module 7, which is a self-calibrating gain. The estimated value of the motor's rotational inertia obtained by the model reference inertia identification algorithm module 7 after correction of gain is used. The desired rotor angular velocity ω output by module 8 is the desired rotational speed setting. ref The rotor angular velocity ω output by differential module 3 m The input is fed into the disturbance-constrained enhanced active disturbance rejection controller module 9, and the initial desired q-axis current i is output after calculation by the disturbance-constrained enhanced active disturbance rejection controller module 9.q * The desired q-axis current i is obtained through the limiting module 10. qsat * The desired q-axis current i output by the limiting module 10 is... qsat * The actual q-axis current i output by Park coordinate transformation module 6 q By subtraction, the q-axis current deviation e is obtained. iq , the q-axis current deviation e iq The input is fed into PI module 12 to obtain the desired q-axis voltage u. q Similarly, the i output by the d-axis desired current module 11 will be... d The actual d-axis current i output by Park coordinate transformation module 6 d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into PI module 13 to obtain the desired d-axis voltage u. d ; the desired voltage u along the d-axis d q-axis desired voltage u q and rotor electrical angle θ e The input is fed into the inverse Park coordinate transformation module 14 to obtain the desired voltage u in the two-phase stationary coordinate system. α u β The desired voltage u obtained α u β The input is sent to the space vector pulse width modulation module 15 to obtain the switching signal S. A S B S C The switching signal is input to the three-phase inverter module 16 to generate a pulse signal, which in turn controls the speed and torque of the permanent magnet motor 1.

[0064] To verify the control performance of the disturbance rejection control method for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement, according to... Figure 1 The system block diagram shown was used to build a system simulation model in Matlab / Simulink, and simulation analysis was performed.

[0065] To verify the superior performance of the disturbance error constraint-enhanced active disturbance rejection controller, the system performance of the velocity loop based on the PI controller, the traditional linear active disturbance rejection controller (LADRC), the disturbance error constraint-enhanced active disturbance rejection controller (DADRC), and the disturbance error constraint-enhanced active disturbance rejection controller (DSADRC) of the present invention was compared and analyzed. Figure 4The figures show a comparison of the speed loop waveforms of the propulsion systems based on the four controllers described above, as illustrated in this embodiment of the invention. The simulation conditions for all four controller systems were set as follows: the desired speed was set to the rated speed of 500 rpm / min, a 7 N·m load was suddenly applied at t=0.5s, and a 7 N·m load was suddenly deflected at t=1.0s. Figure 4 The speed response waveforms show that the speed response times of the three active disturbance rejection controllers are basically the same, and faster than the PI controller (0.06s), at 0.03s. Under sudden load increases, the invented DSADRC speed drop is smaller, at 23 rpm, significantly less than DADRC's 51 rpm, LADRC's 86 rpm, and the PI controller's 106 rpm. Under sudden load decreases, the invented DSADRC speed rise is smaller, at 26 rpm, significantly less than DADRC's 53 rpm, LADRC's 87.5 rpm, and the PI controller's 106 rpm. Therefore, the invented disturbance error constraint-enhanced active disturbance rejection controller has stronger dynamic response capabilities than the PI controller and stronger disturbance rejection capabilities than the PI, LADRC, and DADRC controllers.

[0066] To verify the superiority of the model reference adaptive inertia identification algorithm using self-calibrating gain, simulations were built for both the traditional model reference adaptive inertia identification algorithm and the invented self-calibrating gain model reference adaptive inertia identification algorithm. The simulation conditions were set as follows: the initial value of rotational inertia identification was 0.5 J, the desired rotational speed was the rated speed of 500 rpm / min, a 7 N·m load was suddenly applied at t=0.5 s, and a 7 N·m load was suddenly deducted at t=1.0 s. The rotational inertia identification results of the traditional model reference adaptive inertia identification algorithm with adaptive coefficients q=0.002 and q=0.022, and the self-calibrating gain model reference adaptive inertia identification algorithm are shown in the figure below. Figure 5 As shown. From Figure 5 It can be seen that the identification speed of the invented identification algorithm is basically the same as that of the traditional algorithm using a larger adaptive coefficient q. However, when using a smaller adaptive coefficient q, the identification speed of the invented algorithm is significantly faster than that of the traditional algorithm. Under sudden increases in load, the identification results of the invented algorithm change only slightly, by 0.02 × 10⁻⁶. - 4 kg.m 2 The traditional algorithm, which uses a larger adaptive coefficient q, shows a significant variation in identification results, with a value of 0.27 × 10⁻⁶. -4 kg.m 2 In other words, the invented algorithm has a fast identification speed, and its identification results are less coupled with load changes, thus possessing high identification accuracy.

[0067] To verify that the disturbance error constraint-enhanced active disturbance rejection control method based on inertia identification can solve the problem of control gain mismatch and subsequent performance degradation caused by motor rotational inertia mismatch, the method is as follows: Figure 1 The control system block diagram shown is used to build an overall simulation. The simulation conditions are: the desired rotational speed is a square wave with an amplitude of 500 rpm / min to 250 rpm / min and a period of 1 s, the initial inertia of the controller is set to 0.3 J, and the inertia identification algorithm is started at 3 s. Figure 6 This is a speed tracking diagram of the permanent magnet propulsion motor anti-disturbance control system based on inertia identification and disturbance error constraint enhancement in an embodiment of the present invention when the speed command is a square wave. Figure 7 This is a diagram showing the results of the moment of inertia identification at this point. Observe. Figure 6 , Figure 7 It can be seen that as the inertia identification algorithm starts at 3s, the rotational inertia fed back into the controller approaches the true value. When the speed is suddenly increased, the speed overshoot of the system decreases, and when the speed is suddenly decreased, the speed drop decreases. This indicates that the system performance gradually recovers. That is, the invented method can solve the problem of the decline in system control performance caused by the mismatch of control gain due to inertia mismatch.

[0068] In summary, the disturbance rejection control method for permanent magnet propulsion motors based on inertia identification and disturbance error constraint enhancement invented in this paper enables the motor speed control system to quickly achieve overshoot-free speed response, has stronger robustness to load disturbances, and can address the problem of large control gain error caused by inertia mismatch leading to a decrease in system control performance.

[0069] The detailed descriptions listed above are merely specific descriptions of feasible embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. All equivalent methods or modifications that do not depart from the technology of the present invention should be included within the scope of protection of the present invention.

Claims

1. A disturbance rejection control system for a permanent magnet propulsion motor based on inertia identification and disturbance error constraint enhancement, characterized in that, include: Permanent magnet motor (1), encoder module (2), differential module (3), current sampling module (4), Clark coordinate transformation module (5), Park coordinate transformation module (6), self-correcting gain model reference inertia identification algorithm module (7), desired speed setting module (8), disturbance error constraint enhanced active disturbance rejection controller module (9), limiting module (10), d-axis desired current setting module (11), first PI module (12), second PI module (13), inverse Park coordinate transformation module (14), space vector pulse width modulation module (15), and inverter module (16); The encoder module (2) and the current sampling module (4) acquire the rotor mechanical angle θ of the permanent magnet motor (1). m and three-phase current i a i b i c The rotor mechanical angle θ m The rotor angular velocity ω is obtained through the differential module (3). m Rotor mechanical angle θ m and the number of pole pairs p of the motor n Multiplication yields the rotor electrical angle θ e Three-phase current i a i b i c The current i in the two-phase stationary coordinate system is obtained through the Clark coordinate transformation module (5). α i β Current i in a two-phase stationary coordinate system α i β and rotor electrical angle θ e The input to the Park coordinate transformation module (6) yields the quadrature and direct axis current signals i. q i d The quadrature axis current i q With torque coefficient K t Multiplication yields the electromagnetic torque T of the motor. e The rotor angular velocity ω obtained by the differential module (3) m and the electromagnetic torque T of the motor e The estimated value of the motor's rotational inertia is obtained by inputting it into the self-calibrating gain model reference inertia identification algorithm module (7). Estimated value of the moment of inertia of the motor The desired rotor angular velocity ω output by the desired rotational speed setting module (8) ref and the rotor angular velocity ω output by the differential module (3) m The input is fed into the disturbance error constraint enhanced active disturbance rejection controller module (9), and the disturbance error constraint enhanced active disturbance rejection controller module (9) calculates and outputs the initial desired q-axis current i. q * Initial expected current i q * The desired q-axis current i is obtained through the limiting module (10). qsat * q-axis desired current i qsat * The actual q-axis current i output by the Park coordinate transformation module (6) q By subtracting, we obtain the q-axis current deviation e. iq , the q-axis current deviation e iq The input is fed into the first PI module (12) to obtain the desired q-axis voltage u. q Meanwhile, the i output of the d-axis desired current module (11) d * The actual d-axis current i output by the Park coordinate transformation module (6) d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into the second PI module (13) to obtain the desired d-axis voltage u. d d-axis desired voltage u d q-axis desired voltage u q Rotor electrical angle θ e The input is fed into the inverse Park coordinate transformation module (14) to obtain the desired voltage u in the two-phase stationary coordinate system. α u β The desired voltage u α u β The input to the space vector pulse width modulation module (15) obtains the switching signal S used to drive the inverter module (16). A S B S C The switching signal is input to the three-phase inverter module (16) to generate the motor pulse signal, thereby realizing the closed-loop control of the permanent magnet motor (1).

2. The anti-disturbance control system for permanent magnet propulsion motors according to claim 1, characterized in that, The self-calibrating gain model reference inertia identification algorithm module (7) includes: a discretization processing module (17), a reference model module (18), an adjustable model module (19), a rotational inertia identification sub-model module (20), a first delayed sampling module (21), a self-calibration mechanism module (22), a second delayed sampling module (23), and a continuousization module (24). The discretization module (17) receives the rotor angular velocity ω m and the electromagnetic torque T of the motor e The electromagnetic torque T at time k-1 is obtained after discretization. e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the angular velocity ω of the motor rotor at time k-1. m Rotor angular velocities ω at time (k-1) and (k-2) m (k-2) and input it into the reference model module (18); The reference model module (18) processes the input to obtain the rotor angular velocity ω at time k. m (k); The adjustable model module (19) receives the electromagnetic torque T at time k-1 output by the discretization processing module (17). e The electromagnetic torque T at times (k-1) and (k-2) e (k-2), the rotor angular velocity ω at time k-1 m (k-1), the rotor angular velocity ω at time k-2 m (k-2) and the identification values ​​of the identification parameters output by the second delayed sampling module (23) at time k-1. (k-1), to obtain the estimated rotor angular velocity at time k. ; The rotational inertia identification sub-model module (20) receives ω m (k) T e (k-1), T e (k-2) The identification value at time k is calculated using the adaptive coefficient q(k) output by the self-correction mechanism module (22) at time k (k-1) and the self-correction mechanism module (22). (k) and the identified value of the moment of inertia at time k ; The first delayed sampling module (21) identifies the moment of inertia at time k. Delay processing ; The self-calibration mechanism module (22) receives and The adaptive coefficient q(k) at time k is obtained through processing, and its expression is: The second delayed sampling module (23) will input the k-th time-time identification value. (k) delay is (k-1); The continuous processing module (24) identifies the rotational inertia value at time k. Processed as an estimate of the moment of inertia .

3. The anti-disturbance control system for permanent magnet propulsion motors according to claim 2, characterized in that, The rotor angular velocity ω in the reference model module (18) m The formula for calculating (k) is: In the formula, T s Where is the sampling time, J is the motor's moment of inertia, and h is the identification parameter.

4. The anti-disturbance control system for permanent magnet propulsion motors according to claim 3, characterized in that, The rotor angular velocity estimate at time k in the adjustable model module (19) The calculation formula is: 。 5. The anti-disturbance control system for permanent magnet propulsion motors according to claim 4, characterized in that, The identified value at time k in the rotational inertia identification sub-model module (20) (k) and the identified value of the moment of inertia at time k The calculation formula is: 。 6. The anti-disturbance control system for permanent magnet propulsion motor according to claim 5, characterized in that, The formula for calculating the adaptive coefficient q(k) at time k in the self-correction mechanism module (22) is as follows: In the formula, e J (k) represents the inertia identification velocity at time k, v is the correction intensity coefficient, and q0 is the initial adaptive coefficient.

7. The anti-disturbance control system for permanent magnet propulsion motors according to claim 6, characterized in that, The disturbance error constraint enhanced active disturbance rejection controller module (9) includes: a first-level disturbance error constraint extended state observer module (25), a second-level disturbance error constraint extended state observer module (26), a proportional control law module (27), a control gain calculation module (28), and an initial desired current calculation module (29). The first-stage disturbance error constraint expansion state observer module (25) receives the rotor angular velocity ω output by the differential module (3). m The control quantity u2 obtained from the control quantity calculation and compensation loop is used to observe the rotor angular velocity Z1 and disturbance Z2 of the first-stage observer. The second-level disturbance error constraint extended state observer module (26) receives the disturbance observation value Z2 observed by the first-level disturbance error constraint extended state observer module (25) and the rotor angular velocity ω output by the differential module (3). m In addition to the control quantity u2 obtained from the control quantity calculation and compensation process, the disturbance observation value S2 of the second-level observer is observed. The proportional control law module (27) receives the speed tracking deviation e. s The speed tracking deviation e s The desired rotor angular velocity ω output by the desired rotational speed setting module (8) ref The proportional control law module (27) obtains the result by subtracting Z1 from the first-level disturbance error constraint extended state observer module (25) based on the received e. s Calculate the control quantity u1; In the control quantity calculation and compensation process, the total observed disturbance is obtained by first adding the disturbance observation value Z2 of the first-level observer output by the first-level disturbance error constraint extended state observer module (25) and the disturbance observation value S2 of the second-level observer output by the second-level disturbance error constraint extended state observer module (26). Then the control quantity u1 calculated by the proportional control law module (27) is compared with the total observed disturbance. The difference is used to obtain the control quantity u2 after disturbance compensation; The control gain calculation module (28) receives the estimated moment of inertia obtained by the continuous processing module (24). Calculate the control gain b1 of the controlled system; The initial desired current calculation module (29) receives the control gain b1 and the control quantity u2, and calculates the initial desired current i. q * .

8. The anti-disturbance control system for a permanent magnet propulsion motor according to claim 7, characterized in that, The specific expression for the first-level perturbation error constraint extended state observer is as follows: In the formula, Z1 is the observed rotor angular velocity of the first-stage observer, Z2 is the observed disturbance value of the first-stage observer, and E Z1 E represents the velocity observation error of the first-stage observer. Z2 l1 represents the perturbation observation error of the first-stage observer, and l2 represents the observer coefficients, which are greater than 0, where l1 = ω. 01 l2=ω 01 ω 01 This is the bandwidth of the first-stage observer; The specific expression for the second-level perturbation error constraint extended state observer is as follows: In the formula, S1 is the observed rotor angular velocity of the second-stage observer, S2 is the observed disturbance value of the second-stage observer, and E S1 E represents the velocity observation error of the second-stage observer. S2 l3 represents the perturbation observation error of the second-stage observer, and l4 represents the observer coefficients, which are greater than 0, where l3 = ω 02 l4=ω 02 ω 02 This is the bandwidth of the second-stage observer.

9. The anti-disturbance control system for a permanent magnet propulsion motor according to claim 8, characterized in that, The proportional control law module (27) calculates the expression for the control quantity u1 as follows: In the formula, e s ω represents the speed tracking deviation. ref k represents the desired rotor angular velocity. p Indicates the proportional feedback coefficient; The expression for calculating the control gain b1 by the control gain calculation module (28) is as follows: In the formula, K t p represents the torque coefficient of a permanent magnet motor. n ψ represents the number of pole pairs of the motor. f This refers to the permanent magnet flux linkage of the motor; The initial desired current calculation module (29) calculates the initial desired current i q * The expression is: 。 10. The control method of the permanent magnet propulsion motor anti-disturbance control system according to claim 1, characterized in that, Step 1: Acquire the rotor mechanical angle θ of the permanent magnet motor through the encoder module and the current sampling module. m and three-phase current i a i b i c The rotor mechanical angle θ m The rotor angular velocity ω is then obtained through the differential module. m The rotor mechanical angle θ m and the number of pole pairs p of the motor n Multiplication yields the rotor electrical angle θ e ; the three-phase current i a i b i c The current i in the two-phase stationary coordinate system is obtained using the Clark coordinate transformation module. α i β Then, the current i in the two-phase stationary coordinate system α i β and rotor electrical angle θ e The signal is input into the Park coordinate transformation module to obtain the quadrature-axis current signal i. q Direct-axis current signal i d ; Step 2: Convert the quadrature-axis current i obtained by the Park coordinate transformation module. q With torque coefficient K t Multiplication yields the electromagnetic torque T of the motor. e Then the rotor angular velocity ω obtained by the differential module m and the electromagnetic torque T of the motor e The estimated value of the motor's rotational inertia is obtained by inputting it into the self-calibrating gain model reference inertia identification algorithm module. ; Step 3: The estimated value of the motor's rotational inertia obtained by the self-calibrating gain model reference inertia identification algorithm module. The desired rotor angular velocity ω output by the desired speed setting module ref The rotor angular velocity ω output by the differential module m The input is fed into the disturbance error constraint-enhanced active disturbance rejection controller module to calculate the initial desired output q-axis current i. q * The desired q-axis current i is then obtained through a limiting module. qsat * ; Step 4: Calculate the desired q-axis current i output by the limiting module. qsat * The actual q-axis current i output by the Park coordinate transformation module q By subtraction, the q-axis current deviation e is obtained. iq , the q-axis current deviation e iq The input is fed into the first PI module to obtain the desired q-axis voltage u. q Simultaneously, the i output by the d-axis desired current setting module is... d * The actual d-axis current i output by the Park coordinate transformation module d By subtraction, the d-axis current deviation e is obtained. id d-axis current deviation e id The input is fed into the second PI module to obtain the desired d-axis voltage u. d ; Step 5: Set the desired d-axis voltage u d q-axis desired voltage u q and rotor electrical angle θ e The input is fed into the inverse Park coordinate transformation module to obtain the desired voltage u in the two-phase stationary coordinate system. α u β and u α u β As input to the space vector pulse width modulation module, the switching signal S used to drive the inverter module is obtained. A S B S C The inverter module receives the switching signal S. A S B S C Then, the drive signal of the motor is obtained, which enables the permanent magnet motor to work, forming a closed-loop control.

Images