Model-free predictive current control method based on improved observer

Abstract

The invention discloses a model-free predictive current control method based on an improved observer, and belongs to the technical field of permanent magnet synchronous motor control. The control method comprises the following steps: firstly, by analyzing the frequency characteristic of ESO, selecting an extended state observer order for obtaining balance between the anti-interference capability and the noise suppression capability; secondly, designing a model-free predictive current loop controller, and calculating the observer gain through a bandwidth method; then, a judgment mechanism of a current loop controller is constructed, the mechanism judges whether the system is in a steady state or not at the moment according to the current deviation value in the current state, and therefore the observer order with the better control effect is selected. According to the model-free predictive current control method based on the improved observer, the anti-interference performance of the system and the noise suppression capability of the current are remarkably improved, it is guaranteed that the improved observer has rapid tracking response and excellent steady-state performance, and the robustness of the control system is enhanced.

Description

Technical Field

[0001] This invention belongs to the field of permanent magnet synchronous motor control technology, specifically relating to a model-free predictive current control method based on an improved observer. Background Technology

[0002] To address the performance degradation issues of traditional control methods for permanent magnet synchronous motors (PMSMs) under parameter mismatch and complex operating conditions, Model-Free Predictive Current Control (MFCC) has attracted significant attention as a novel control approach. The core idea of ​​MFCC is to estimate the motor's operating state in real time through data-driven methods, thereby reducing reliance on motor parameters. An Extended State Observer (ESO) is an observer used to estimate the system's internal state and external disturbances, providing real-time feedback to the controller to improve system robustness. Introducing an ESO into MFCC can further improve the accuracy of current prediction, thus enhancing the stability of the control system.

[0003] Conventional second-order ESOs have relatively low observation accuracy and insufficient robustness against complex nonlinear disturbances and higher-order uncertainties. By extending the state dimension, higher-order ESOs can more effectively handle higher-order nonlinear disturbances and uncertainties in the system, exhibiting stronger robustness. However, higher-order ESOs are more sensitive to measurement noise and are prone to introducing high-frequency noise, which reduces the noise suppression capability of the current. Summary of the Invention

[0004] To address the issues of low observation accuracy and poor anti-interference capability of conventional second-order ESOs, and the reduced noise suppression capability of current due to the sensitivity of higher-order ESOs to measurement noise, this method designs an improved hybrid observer in the inner current loop based on the traditional observer. By judging the real-time state of the system, the ESO of the different orders can be switched autonomously, so that the system can have both good current noise suppression capability and anti-interference capability.

[0005] To achieve the above objectives, this invention proposes a model-free predictive current control method based on an improved observer, the specific implementation steps of which are as follows:

[0006] Step (1): Establish the hyperlocal mathematical model of the permanent magnet synchronous motor (PMSM) based on the current dynamic equation of the PMSM, and derive the stator voltage reference equation.

[0007] Step (2): Based on the hyperlocal mathematical model, construct the extended state second-order linear observer model and design the observer gain of the extended state second-order linear observer ESO.

[0008] Step (3): Extend the second-order linear observer model to ESOs of different orders, and determine the transfer function from the observed value to the total disturbance based on the ESO of different orders. and the error transfer function of disturbance estimation By analyzing the frequency characteristics of ESOs of different orders, a fourth-order linear ESO and a second-order linear ESO were selected.

[0009] Step (4): Design a model-free predictive current loop controller based on a fourth-order observer ESO, and design the observer parameters by configuring poles using the bandwidth method, and construct the gain expression of the fourth-order observer ESO.

[0010] Step (5): Construct an improved observer ESO current loop control mechanism to autonomously switch the order of the ESO to second or fourth order according to the system state.

[0011] A model-free predictive current control device based on an improved observer, comprising:

[0012] Mathematical model building module: Based on the current dynamic equation of the permanent magnet synchronous motor, a hyperlocal mathematical model of the permanent magnet synchronous motor (PMSM) is established, and the stator voltage reference equation is derived.

[0013] Observer model design module: Based on the hyperlocal mathematical model, construct the extended state second-order linear observer model, and design the observer gain of the extended state second-order linear observer ESO.

[0014] Observer selection module: Based on the second-order linear observer model, this module extends to observers of different orders (ESOs), and selects the total disturbance from the observed value according to the transfer function of the ESO for each order. and the error transfer function of disturbance estimation By analyzing the frequency characteristics of ESOs of different orders, a fourth-order linear ESO and a second-order linear ESO were selected.

[0015] Gain Expression Construction Module: Design a model-free predictive current loop controller based on a fourth-order observer ESO, and design the observer parameters by configuring poles using the bandwidth method to construct the gain expression of the fourth-order observer ESO.

[0016] Switching Module: Construct an improved observer ESO current loop control mechanism to autonomously switch the order of the ESO to second or fourth order based on the system state.

[0017] An electronic device includes: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method.

[0018] A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to implement the method described thereon.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] This method overcomes the limitations and shortcomings of traditional extended state observers in current tracking, enhancing the robustness of the control system. It improves control performance by modifying the controller structure without adding new hardware. Furthermore, the invention is clear in its concept and structure, making it easy to implement in engineering. Attached Figure Description

[0021] Figure 1 This is a structural block diagram of a model-free predictive current control method based on an improved observer according to the present invention;

[0022] Figure 2 The diagram shows a comparative experimental waveform of the quadrature-axis current iq under the condition of parameter mismatch: conventional model-free predictive current control, model-free predictive current control based on second-order ESO, model-free predictive current control based on fourth-order ESO, and model-free predictive current control based on improved ESO proposed in this invention.

[0023] Figure 3 The waveforms show a comparison of three-phase current ia under conventional model-free predictive current control, model-free predictive current control based on second-order ESO, model-free predictive current control based on fourth-order ESO, and model-free predictive current control based on improved ESO proposed in this invention.

[0024] Figure 4 The experimental waveforms comparing the rotational speed n of the motor under parameter mismatch conditions are: conventional model-free predictive current control, model-free predictive current control based on second-order ESO, model-free predictive current control based on fourth-order ESO, and model-free predictive current control based on improved ESO proposed in this invention. 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 merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. To achieve the above objectives, this invention adopts the following technical solution.

[0026] As attached Figure 1The diagram shows a control block diagram of a model-free predictive current control method based on an improved observer, including a speed loop, a current loop, and a switchable improved ESO controller. The improved ESO current controller can autonomously switch between different orders according to the real-time state of the system to achieve good steady-state performance and anti-interference capability. The specific implementation steps are as follows:

[0027] First, based on the current dynamic equation of the permanent magnet synchronous motor, the voltage equation of the permanent magnet synchronous motor in the dq rotating coordinate system is:

[0028] ;

[0029] Furthermore, the dynamic equation for the current can be expressed as:

[0030] ;

[0031] in, and These are the voltages along the d-axis and q-axis, respectively. and These are the d-axis and q-axis currents, respectively. The resistance of the stator winding, and The inductances along the d-axis and q-axis are respectively. The rotor angular velocity, It is the magnetic flux of a permanent magnet.

[0032] A hyperlocal mathematical model of a permanent magnet synchronous motor (PMSM) is established. The internal parameters of the motor are... , and The changes in [variable name] and the external disturbances it experiences are defined as the total disturbance. The mathematical model expression for the hyperlocal model is:

[0033] ;

[0034] in, It is the stator current. It is the stator voltage. It is a stator inductor. It is the stator resistance. It is the total disturbance of the system (including external load disturbances, parameter changes, etc.).

[0035] Construct a conventional second-order linear extended state observer structure, using stator current and the total disturbance of the system The second-order linear ESO model with state variables is as follows:

[0036] ;

[0037] in, This is an estimated value for the stator current. It is an estimate of the total system disturbance. and This is the observer gain of ESO.

[0038] The observer gain is selected for the second-order linear ESO. Observer gain and The pole placement method can be used to determine this, that is, to obtain it through the closed-loop eigenvalues ​​of a second-order extended state observer:

[0039] ;

[0040] The poles are uniformly configured at Place, that is ,in Let be the observer bandwidth. Then the observer gain is... and They are respectively:

[0041] ;

[0042] By analyzing the frequency characteristics of ESOs of different orders, an observer order that strikes a balance between anti-interference capability and noise suppression capability is selected. The error dynamic equation of the extended state observer is:

[0043] ;

[0044] in, For extended order, For the gain of the extended state observer, For the first One error state variable, For the first The derivative of each error state variable, It represents the nth derivative of the system disturbance.

[0045] By employing a bandwidth tuning method, the closed-loop poles of the observer are set at the same location. Then the gain coefficient will satisfy:

[0046] ;

[0047] Therefore, the characteristic polynomial is:

[0048] ;

[0049] Based on the aforementioned characteristic polynomial, the transfer functions from the observed values ​​to the total disturbance for observers of different orders are derived. and the error transfer function of disturbance estimation Through analysis, the transfer function from the observed values ​​to the total disturbance can be obtained:

[0050] ;

[0051] The error transfer function for disturbance estimation is then:

[0052] ;

[0053] Analyzing the transfer functions of both, it is found that as the ESO order increases, the anti-interference capability improves due to more accurate estimation and faster dynamic response, but the noise suppression capability weakens due to increased high-frequency gain and sensitivity to complex structures. This is based on the extended state observers of different orders. and Based on the frequency characteristics, this invention employs second-order and fourth-order ESOs to achieve a superior switching control effect.

[0054] The observer gain of a second-order linear ESO is calculated, and the observer gain is designed by configuring poles using the bandwidth method. , , and That is, it is obtained through the closed-loop eigenvalues ​​of the fourth-order extended state observer:

[0055] ;

[0056] The poles are uniformly configured at Place, that is ,in Let be the observer bandwidth. Then the observer gain is... , , and They are respectively:

[0057] ;

[0058] Design an improved ESO current loop control mechanism and construct a model-free predictive current loop controller based on this improved mechanism. First, the given current is... and the current fed back through the current loop The absolute value of the difference is defined as the current error, denoted by the symbol. Represented as:

[0059] ;

[0060] This error signal The model-free predictive controller in the current loop determines the appropriate ESO order for the given state and feeds this information back to the controller to optimize the control input.

[0061] This judgment mechanism needs to set a threshold. This threshold reflects the magnitude of the current value deviation in the current state and can be used to determine whether the system is in a steady state. If Greater than the threshold If the system is in a transient state, a fourth-order ESO with stronger anti-interference capabilities will be used. If Less than the threshold When the system is in a steady state, the observer will switch to a second-order ESO. When affected by a sinusoidal disturbance and the function is... The maximum output steady-state error is derived as follows:

[0062] ;

[0063] in, It is the signal frequency ω and the observer bandwidth The ratio of, where This represents the bandwidth of the closed-loop system.

[0064] Therefore, the error at different frequencies can be calculated using the above formula, and the threshold should be set to the maximum value of the steady-state error:

[0065] ;

[0066] in, The exponent represents the frequency components involved in the interference signal. The improved current loop controller can autonomously switch between different orders of ESO according to the system status, achieving better current control without affecting anti-interference performance and noise suppression capability.

[0067] By comparison Figure 2 , Figure 3 and Figure 4 The waveforms of the experimental results can verify the effectiveness of the present invention. Figure 2 , Figure 3 and Figure 4 The figures show experimental comparison waveforms of quadrature-axis current iq, three-phase current ia, and speed n for conventional model-free predictive current control of the motor under parameter mismatch conditions, model-free predictive current control based on second-order ESO, model-free predictive current control based on fourth-order ESO, and model-free predictive current control based on improved ESO proposed in this invention.

[0068] To ensure a fair comparison of the control performance of each algorithm, the parameters of the speed loop controller and the initial conditions of the motor were identical across all four algorithms. The initial condition for the experiment was that the motor was running under no-load, and an initial value of 0 was added at 0.35s. The final value is 10 The step load torque. The motor's initial speed is 0 rpm, the given speed is 2000 rpm, and then it decelerates to 1000 rpm in 0.65 seconds. The motor's rated current is 20A, and the observer's bandwidth is set to... The threshold is set to δ = 0.05A. The degree of motor parameter mismatch in the current controller is determined by the inductance... Mismatch 2 times, resistance Mismatch 0.5x, magnetic flux Mismatch 1.5 times.

[0069] Through analysis and comparison Figure 2 , Figure 3 and Figure 4 The waveform diagrams show the dynamic response and steady-state performance of four control algorithms under motor parameter mismatch. Figure 2 The comparison of the cross-axis current iq waveforms shows that the fourth-order ESO is most sensitive to current noise when the system is in steady state. The improved switchable ESO control algorithm with autonomously switchable order proposed in this invention exhibits lower current ripple in steady state than the fourth-order ESO and conventional model-free predictive control, indicating that the proposed method has strong current noise suppression capabilities. Furthermore, the waveforms show that when the system is in transient states of acceleration / deceleration and sudden load increases, the peak current waveform of the switchable ESO control algorithm proposed in this invention is lower than that of the second-order ESO and conventional model-free predictive control, demonstrating good anti-interference capabilities.

[0070] Figure 3 This is a comparison of the three-phase current ia waveforms for four control methods. As can be seen from the graphs, among the four methods, the switchable ESO control algorithm proposed in this invention exhibits the smallest current ripple in steady state, demonstrating stronger current noise suppression compared to the model-free predictive control of a fourth-order ESO. Furthermore, the waveforms show that the peak current waveform of the switchable ESO control algorithm is smaller than that of a second-order ESO and conventional model-free predictive control, indicating better anti-interference capabilities.

[0071] Figure 4 This is a waveform comparison of the speed changes for four control algorithms. As can be seen from the graph, all four control methods can quickly reach the specified speed during motor startup and deceleration. Furthermore, when a sudden load is applied to the motor, all four methods can quickly return to the original speed value. This demonstrates that the control method of this invention has good speed and dynamic response performance compared to the other three control algorithms.

[0072] In summary, the model-free predictive current control method based on an improved observer proposed in this invention achieves a good balance between noise suppression and anti-interference capabilities. It can autonomously switch to an ESO order with better control performance based on the current system state. It possesses superior current noise suppression compared to traditional fourth-order ESOs, and stronger anti-interference capabilities than conventional second-order ESOs. Furthermore, it exhibits excellent dynamic response and steady-state performance.

Claims

1. A model-free predictive current control method based on an improved observer, characterized in that, The steps are as follows: Step (1): Establish the hyperlocal mathematical model of the permanent magnet synchronous motor (PMSM) based on the current dynamic equation of the PMSM, and derive the stator voltage reference equation. Step (2): Based on the hyperlocal mathematical model, construct the extended state second-order linear observer model and design the observer gain of the extended state second-order linear observer ESO. Step (3): Extend the second-order linear observer model to ESOs of different orders, and determine the transfer function from the observed value to the total disturbance based on the ESO of different orders. and the error transfer function of disturbance estimation By analyzing the frequency characteristics of ESOs of different orders, a fourth-order linear ESO and a second-order linear ESO were selected. Step (4): Design a model-free predictive current loop controller based on a fourth-order observer ESO, and design the observer parameters by configuring poles using the bandwidth method, and construct the gain expression of the fourth-order observer ESO. Step (5): Construct an improved observer ESO current loop control mechanism to autonomously switch the order of the ESO to second or fourth order according to the system state.

2. The model-free predictive current control method based on an improved observer according to claim 1, characterized in that, Step (1) includes: In the synchronous rotating coordinate system, i.e., the dq coordinate system, the current dynamic equation of the PMSM is expressed as: ； in, and These are the d-axis and q-axis currents, respectively. The resistance of the stator winding, and The inductances along the d-axis and q-axis are respectively. The rotor angular velocity, and These are the voltages along the d-axis and q-axis, respectively. It is the magnetic flux of a permanent magnet; Define parameters , and The change is the total disturbance The expression for the hyperlocal mathematical model is: ； in, It is the stator current. It is the stator voltage. It is a stator inductor. It is the stator resistance; Based on the hyperlocal mathematical model, the stator voltage reference equation in the synchronous reference is derived: 。 3. The model-free predictive current control method based on an improved observer according to claim 2, characterized in that, Step (2) includes: constructing a hyperlocal mathematical model based on the stator current established in the above steps. Total disturbance A second-order linear ESO model with state variables: ； in, This is an estimated value for the stator current. It is an estimate of the total system disturbance. and This is the observer gain of ESO.

4. The model-free predictive current control method based on an improved observer according to claim 3, characterized in that, For a second-order linear ESO model, the observer gain and The selection adopts the pole placement method, that is, through the system matrix of the state observer. We can obtain the closed-loop eigenvalues ​​by: ； in, For the observer bandwidth, the matrix Observer gain matrix ,matrix So that the characteristic roots all fall within Place, and The following was obtained: 。 5. The model-free predictive current control method based on an improved observer according to claim 4, characterized in that, Step (3) includes: analyzing the transfer function from the observations to the total disturbance: ； The error transfer function for disturbance estimation is: ； Choose either a second-order linear ESO or a fourth-order linear ESO.

6. The model-free predictive current control method based on an improved observer according to claim 5, characterized in that, Step (4) includes: for a fourth-order linear ESO, the observer gain , , and The selection of poles also adopts the pole placement method, that is, through the system matrix of the state observer. Determining the closed-loop eigenvalues: ； in, The observer bandwidth; ensuring that all eigenvalues ​​fall within the range of the observer. Place, , , and The following was obtained: 。 7. The model-free predictive current control method based on an improved observer according to claim 6, characterized in that, Step (5) includes: 1) Define the current error signal : Given current and the current fed back through the current loop The absolute value of the error between them is defined as the sign. ; 2) Judgment mechanism: If Greater than the threshold Then a fourth-order ESO is used; when The absolute value is less than When the observer is affected by a sinusoidal perturbation and the function is... The maximum output steady-state error is derived as follows: ； in, It is the signal frequency ω and the observer bandwidth The ratio of, where Indicates bandwidth; The error at different frequencies is calculated using the above formula, and the threshold is set as the maximum value of the steady-state error: ； in, The index represents the frequency components involved in the interference signal.

8. A model-free predictive current control device based on an improved observer, characterized in that, include: Mathematical model building module: Based on the current dynamic equation of the permanent magnet synchronous motor, a hyperlocal mathematical model of the permanent magnet synchronous motor (PMSM) is established, and the stator voltage reference equation is derived. Observer model design module: Based on the hyperlocal mathematical model, construct the extended state second-order linear observer model, and design the observer gain of the extended state second-order linear observer ESO. Observer selection module: Based on the second-order linear observer model, this module extends to observers of different orders (ESOs), and selects the total disturbance from the observed value according to the transfer function of the ESO for each order. and the error transfer function of disturbance estimation By analyzing the frequency characteristics of ESOs of different orders, a fourth-order linear ESO and a second-order linear ESO were selected. Gain Expression Construction Module: Design a model-free predictive current loop controller based on a fourth-order observer ESO, and design the observer parameters by configuring poles using the bandwidth method to construct the gain expression of the fourth-order observer ESO. Switching Module: Construct an improved observer ESO current loop control mechanism to autonomously switch the order of the ESO to second or fourth order based on the system state.

9. An electronic device, characterized in that, include: One or more processors; A memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, cause the processor to perform the method described in any one of claims 1 to 7.

Images