Multi-target robust model predictive control method for permanent magnet synchronous motor

Abstract

The invention discloses a multi-target robust model predictive control method for a permanent magnet synchronous motor, which is used for a three-level inverter power supply permanent magnet synchronous motor system with a quadruple BOOST booster circuit, and comprises the following steps: firstly, carrying out system modeling and discretization, then carrying out multi-target cost function design, solving a multi-target optimization problem based on compromise programming and a fuzzy decision method, and finally, carrying out multi-target robust model predictive control on a three-level inverter power supply permanent magnet synchronous motor system with a quadruple BOOST booster circuit. And calculating a delay compensation strategy and a robustness enhancement measure, and finally calculating the output voltage of the quadruple Boost circuit and the output voltage of the inverter. The method improves the DC bus voltage, improves the output power and efficiency, enhances the robustness of the system to parameter change and uncertainty, realizes multi-target optimization control, ensures that the system stably and efficiently operates under complex working conditions, and is suitable for the fields of speed regulation of motors with booster circuits and the like.

Description

Technical Field

[0001] The invention belongs to the technical field of power electronic equipment and control thereof, and in particular relates to a multi-objective robust model predictive control method for a permanent magnet synchronous motor. Background Art

[0002] As a potential control strategy, model predictive control has been widely used in the field of power converter control in recent years to improve its control performance. In addition, model predictive control is a control method based on a system dynamic model, which is highly dependent on the accuracy of system parameters. If the parameters are inaccurate, it will directly affect the steady-state tracking performance of the predictive control, causing static errors in the control system and reducing its reliability. In actual operation, the dynamic behavior of the controlled object, the inaccuracy and uncertainty of the model, and the unmeasurable parameters will lead to a mismatch between the actual parameters and the nominal values ​​in the prediction model, thereby affecting the control effect and reducing the robustness of the control.

[0003] The control of permanent magnet synchronous motors needs to take into account the performance requirements of multiple aspects. The existing multi-objective optimization methods rely on experience to select weight coefficients for the coefficients, which leads to a lack of objectivity and consistency, and it is difficult to form a unified standard. At the same time, it also lacks universality and is difficult to deal with uncertainty, resulting in performance degradation in the face of unexpected situations, and it is impossible to ensure that the system can operate stably under various working conditions. Compromise planning and fuzzy decision-making methods can effectively overcome these problems. Summary of the invention

[0004] In order to achieve optimal control of a permanent magnet synchronous motor, the present invention proposes a multi-objective robust model predictive control method for a permanent magnet synchronous motor inverter based on compromise planning and fuzzy decision-making method, which is used in a three-level inverter-powered permanent magnet synchronous motor system with a quadruple BOOST boost circuit to improve the system DC bus voltage and enhance the robustness of the system under parameter changes and uncertainties. A two-step prediction delay is used to compensate for the calculation delay to achieve multi-objective optimal control.

[0005] The technical solution adopted by the present invention to solve the technical problem is: a multi-objective robust model predictive control method for a permanent magnet synchronous motor, which is used in a three-level inverter-powered permanent magnet synchronous motor system with a quadruple BOOST boost circuit, comprising the following steps:

[0006] S1, establish the overall system dynamic model including quadruple boost circuit and permanent magnet synchronous motor, use forward Euler method to discretize the whole system dynamic model, obtain the discrete state space model for subsequent predictive control calculation, and construct a multi-objective cost function including current tracking, neutral point voltage, common mode voltage, switching frequency, quadruple boost circuit output voltage stability and efficiency;

[0007] S2, based on the consideration of the four control objectives of the inverter, namely, current tracking control, neutral point voltage control, common mode voltage control and switching frequency control, the output voltage stability and efficiency of the quadruple boost circuit are considered, and the multi-objective optimization problem is solved based on compromise planning and fuzzy decision-making methods;

[0008] S3, firstly, the solution set on the Pareto front of the multi-objective problem is solved by measuring the difference between the optimal objective function value and the benchmark value, then the satisfaction of the solution on the Pareto front is evaluated by fuzzy decision analysis, and finally the best compromise solution is compared and selected by maximizing the minimum satisfaction principle;

[0009] S4 adopts two-step prediction delay compensation calculation delay and combines dynamic error correction to enhance robustness: predict the system state and related variables at two future moments k+1 and k+2, correct and adjust the dynamic error, and consider the impact of the quadruple Boost circuit on the current when calculating the error between the predicted current and the actual measured current. Correct the error calculation method to make it more accurately reflect the actual error of the system; based on the corrected error, the formula The predicted current is corrected, where σ1 is a weighting factor 0<σ1≤1. The error correction effect is optimized by adjusting the value of σ1, thereby improving the robustness of the system to changes in overall system parameters, including changes in boost circuit parameters and permanent magnet synchronous motor parameters.

[0010] S5, read reference current and extrapolate using the Lagrangian method and Measure the relevant system variables at the current time k, calculate the output voltage of the quadruple boost circuit and the inverter output voltage under the αβ coordinates The modified discrete state space model is used to predict the system state at time k+1 and k+2, based on the neutral point voltage and Calculate the switch state change Get the common mode voltage u from the lookup table cmv (i) Using the compromise planning method, fuzzy decision making and minimum and maximum satisfaction principle, we can select the optimal solution. dcom The minimum switch state combination is used as the optimal control action and is applied to the inverter and quad-Boost circuit respectively to achieve precise control of the entire system.

[0011] Furthermore, in step S1, a differential dynamic model of the permanent magnet synchronous motor in the dq rotating reference frame is established:

[0012]

[0013] in is the stator dq voltage, i dq is the stator dq current, R s is the stator winding resistance, L d and L q is the stator inductance dq, ω e is the electrical rotation frequency, Ψ pm is the permanent magnetic flux, T e and T L are electromagnetic torque and load torque, Z p is the number of extreme pairs, J m is the moment of inertia, B v is the friction coefficient, T s is the sampling interval;

[0014] Electromagnetic Torque of Permanent Magnet Synchronous Motor Where L d =L q =L s ;

[0015] According to the structure and working principle of the three-level inverter with quadruple Boost circuit, in order to describe the relationship between input voltage, inductor current, capacitor voltage and output voltage, the inductor current state equation of the quadruple Boost circuit in continuous current mode is established as follows:

[0016]

[0017] The capacitor voltage state equation is:

[0018] Using the dq current dynamics and the switching state of the three-level inverter, the neutral point voltage dynamic equation is obtained:

[0019]

[0020] Where C = C dc1 =C dc2 is the support capacitance;

[0021] Furthermore, in step S1, the forward Euler method is used to discretize the entire system dynamic model to obtain the following discrete state space model for subsequent predictive control calculation: m (k+1)=A d (k)x m (k)+B d u m (k)+H d (k), where

[0022] A d (k), B d and H d(k) are the system matrix, input matrix and feedback matrix respectively:

[0023]

[0024] Define the neutral point voltage prediction equation:

[0025]

[0026] Furthermore, the current tracking control in step S2 is to calculate the predicted reference current by the Lagrangian method, and construct the cost function term f according to the error between the predicted current and the reference current. i dcom =|i ref (k+1)-i p (k+1)|; Calculate the neutral point voltage control cost function term based on the deviation between the predicted value and the actual value of the neutral point voltage The common-mode voltage control cost function term is constructed based on the difference between the predicted value of the common-mode voltage and the expected limit value. By calculating the change in switch state And weighted, get the switching frequency control cost function term

[0027] Furthermore, in order to ensure that the DC bus voltage is stable within a suitable range and meet the requirements of permanent magnet synchronous motor drive, a cost function term f is added to the output voltage stability of the quadruple Boost circuit. Boost , which is related to the deviation between the predicted value of the boost circuit output voltage and the set reference value: At the same time, the efficiency of the quadruple Boost circuit is considered, and the cost function f related to efficiency is constructed by analyzing the power loss in the circuit (such as inductor resistance loss, switch tube conduction and shutdown loss, etc.). eff , further optimize the overall system performance:

[0028] Furthermore, in the process of executing the multi-objective compromise planning in step S3, other constraints must also be considered:

[0029]

[0030] in is the Chebyshev distance, is the minimum value of the objective function, ω is the weight coefficient, n and N are the current number of iterations and the maximum number of iterations respectively.

[0031] Going a step further, we can search for the best compromise by the following formula:

[0032]

[0033] In the formula is the satisfaction of the objective function for the nth planning problem, f k,min and f k,max is the minimum and maximum value of the Pareto solution of the objective function; then through the formula Compare these satisfaction levels to identify the best compromise and determine the optimal operating state of the equipment, where S n is the minimum value of the satisfaction of the nth planning decision of the objective function, S max is the maximum value among all minimum satisfaction values.

[0034] Furthermore, based on the corrected error, the predicted current is corrected, and the correction formula is: Where σ1 is a weighting factor 0<σ1≤1. By adjusting the value of σ1, the error correction effect is optimized and the robustness of the system to the overall system parameter changes including the boost circuit parameter changes and the permanent magnet synchronous motor parameter changes is improved:

[0035]

[0036] The beneficial effects of the present invention are:

[0037] The present invention achieves significant improvement in multiple aspects of performance by adding a quadruple Boost circuit to the front side of the inverter.

[0038] On the one hand, the quadruple Boost circuit of the present invention effectively improves the DC bus voltage, and it works stably in the inductor current continuous mode. Based on the reasonably derived voltage gain formula and the accurate analysis of the circuit component characteristics, it ensures that the permanent magnet synchronous motor operates at a more suitable voltage, thereby improving the motor output power and efficiency; at the same time, an improved dynamic error correction method is adopted to fully consider the impact of the boost circuit on the system, correct the current error calculation and prediction, enhance the system's robustness to overall parameter changes, reduce parameter change interference, and improve reliability and stability.

[0039] On the other hand, the present invention redesigns the multi-objective cost function, comprehensively considers the output voltage stability, efficiency and original control objectives of the boost circuit, adopts a fuzzy decision-making method to optimize multiple performance indicators, improves the boost circuit efficiency and DC bus voltage stability while ensuring current tracking accuracy and motor performance, and realizes multi-objective optimization control; and calculates the delay compensation strategy in combination with the dynamic characteristics of the boost circuit, uses a two-step prediction time domain method and fuzzy decision-making to determine the improved cost function of the priority coefficient, evaluates and selects all switch state combinations, compensates for calculation delays, reduces control variable oscillations, improves dynamic response performance, and ensures that the system operates stably and efficiently under complex structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1The three-level inverter topology structure with a quadruple BOOST boost circuit of the present invention;

[0041] Figure 2 The block diagram of the multi-objective robust model predictive control strategy for three-level inverters;

[0042] Figure 3 This is the multi-objective robust model predictive control flow chart of the three-level inverter in step five. DETAILED DESCRIPTION

[0043] The present invention is further described in detail with reference to the accompanying drawings and embodiments as follows.

[0044] Reference Figure 1 , Figure 2 and Figure 3 As shown, the present invention discloses a multi-objective robust model predictive control method for an inverter with a quadruple BOOST boost circuit, a multi-objective robust model predictive control method for a permanent magnet synchronous motor powered by a three-level inverter that solves multi-objective optimization by using compromise planning and fuzzy decision-making methods, and in particular, a multi-objective robust model predictive control method for a permanent magnet synchronous motor driven by an inverter with a quadruple boost circuit, which is suitable for applications in the fields of motor speed regulation with a boost circuit, etc., firstly, a relevant dynamic model is established and discretized, and a multi-objective cost function including current tracking, neutral point voltage, common mode voltage, switching frequency control, and output voltage stability and efficiency of a quadruple Boost boost circuit is designed, and a compromise planning and fuzzy decision-making method is used to solve the multi-objective optimization problem, a two-step prediction delay compensation strategy is added, and a dynamic error correction method is used to enhance robustness, so as to comprehensively improve the control performance, multi-objective optimization capability, robustness and adaptability to calculation delay of the system under complex working conditions. A multi-objective robust model predictive control method for an inverter with a quadruple BOOST boost circuit is proposed. The relevant dynamic model is established and discretized. A multi-objective cost function including current tracking, neutral point voltage, common mode voltage, switching frequency control, output voltage stability and efficiency of the quadruple Boost boost circuit is designed. The multi-objective optimization problem is solved by using compromise planning and fuzzy decision-making methods. A two-step prediction delay compensation strategy is added. The dynamic error correction method is used to enhance robustness. The control performance, multi-objective optimization capability, robustness and adaptability to calculation delay of the system under complex working conditions are comprehensively improved. This is achieved through the following steps.

[0045] Step 1: System modeling and discretization.

[0046] Establish an overall system dynamic model including quadruple Boost circuit and permanent magnet synchronous motor.

[0047] The forward Euler method is used to discretize the dynamic model of the entire system to obtain a discrete state space model for subsequent predictive control calculations. A multi-objective cost function including current tracking, neutral point voltage, common mode voltage, switching frequency, output voltage stability and efficiency of the quad-Boost circuit is constructed.

[0048] For the permanent magnet synchronous motor part, the differential dynamic model is established in the dq rotating reference frame:

[0049]

[0050] in is the stator dq voltage, i dq is the stator dq current, R s is the stator winding resistance, L d and L q is the stator inductance dq, ω e is the electrical rotation frequency, Ψ pm is the permanent magnetic flux, T e and T L are electromagnetic torque and load torque, Z p is the number of extreme pairs, J m is the moment of inertia, B v is the friction coefficient, T s is the sampling interval.

[0051] Electromagnetic Torque of Permanent Magnet Synchronous Motor Where L d =L q =L s .

[0052] According to the structure and working principle of the three-level inverter with quadruple Boost circuit, the state equation of the quadruple Boost circuit in continuous current mode is established to describe the relationship among input voltage, inductor current, capacitor voltage and output voltage.

[0053] The inductor current state equation is:

[0054]

[0055] The capacitor voltage state equation is:

[0056] Using the dq current dynamics and the switching state of the three-level inverter, the neutral point voltage dynamic equation is obtained:

[0057]

[0058] Where C = C dc1 =C dc2 is the support capacitor.

[0059] The three-level inverter is composed of three identical branches, each of which is composed of four switches and can be described by three switch states: 1, 0 and -1.

[0060] Therefore, there are 27 possible combinations of switch states corresponding to the 19 inverter output voltages obtained in the αβ-coordinates. The relationship between the three-level inverter output voltage and the switch state is:

[0061]

[0062] The common mode voltage expression is: where u AZ Defined as the voltage between the phase (A) and the neutral point voltage (Z).

[0063] Then, the forward Euler method is used to discretize the entire system dynamic model to obtain a discrete state space model for subsequent predictive control calculations: m (k+1)=A d (k)x m (k)+B d u m (k)+H d (k), where

[0064] A d (k), B d and H d (k) are the system matrix, input matrix and feedback matrix respectively:

[0065]

[0066] Define the neutral point voltage prediction equation:

[0067]

[0068] The state variable x(k) includes the stator current and speed of the motor, the inductor current and capacitor voltage of the boost circuit, etc. The input variable u(k) can be the switch state control signal of the inverter, etc. Through discretization, the continuous dynamic model is converted into a discrete model suitable for digital calculation, so as to perform subsequent predictive control calculations in the digital controller.

[0069] Step 2: Multi-objective cost function design.

[0070] A multi-objective cost function is designed. On the basis of considering the four control objectives of the inverter, namely current tracking control, neutral point voltage control, common mode voltage control and switching frequency control, the output voltage stability and efficiency of the quadruple Boost circuit are added into consideration. The multi-objective optimization problem is solved based on compromise planning and fuzzy decision-making methods.

[0071] For current tracking control, the predicted reference current is calculated by the Lagrangian method, and the cost function term f is constructed based on the error between the predicted current and the reference current. i dcom =|i ref (k+1)-i p (k+1)|.

[0072] The neutral point voltage control cost function term is calculated based on the deviation between the predicted value and the actual value of the neutral point voltage.

[0073] The common-mode voltage control cost function term is constructed based on the difference between the predicted value of the common-mode voltage and the expected limit value.

[0074] By calculating the change in switch state And weighted, get the switching frequency control cost function term

[0075] In order to ensure that the DC bus voltage is stable within a suitable range and meet the drive requirements of the permanent magnet synchronous motor, a cost function term f is added to the output voltage stability of the boost circuit. Boost , which is related to the deviation between the predicted value of the boost circuit output voltage and the set reference value:

[0076] At the same time, considering the efficiency of the boost circuit, by analyzing the power loss in the circuit (such as inductor resistance loss, switch tube conduction and shutdown loss, etc.), a cost function f related to efficiency is constructed. eff , further optimize the overall system performance:

[0077] Step three, after solving the multi-objective optimization problem based on compromise programming and fuzzy decision-making methods, solve the Pareto frontier solution set.

[0078] Firstly, the solution set on the Pareto front of the multi-objective problem is solved by measuring the difference between the optimal objective function value and the benchmark value. The constraints need to be considered in the solution process. Then, fuzzy decision analysis is used to evaluate the satisfaction of these solutions on the Pareto front. Finally, the principle of maximizing the minimum satisfaction is used to compare and select the best compromise solution.

[0079] In performing this multi-objective trade-off planning, other constraints must be considered:

[0080]

[0081] in is the Chebyshev distance, is the minimum value of the objective function, ω is the weight coefficient, n and N are the current number of iterations and the maximum number of iterations respectively.

[0082] After solving the Pareto frontier solution, the next step is to normalize the objective function values ​​and scale them to the range of [0,1] to more effectively analyze and compare the results. To this end, this study uses fuzzy decision analysis technology to evaluate the degree to which each Pareto solution satisfies different objective functions and further searches for the best compromise solution. The specific method is as follows:

[0083]

[0084] In the formula is the satisfaction of the objective function for the nth planning problem, f k,min and f k,max is the minimum and maximum value of the Pareto solution of the objective function.

[0085] After obtaining the satisfaction degree of the objective function relative to the Pareto solution using the fuzzy decision-making method, a specific formula is used to compare these satisfaction degrees to identify the optimal compromise solution and determine the optimal operating state of the equipment at the same time: Where S n is the minimum value of the satisfaction of the nth planning decision of the objective function, S max is the maximum value among all minimum satisfaction values.

[0086] Calculate the satisfaction of the k-th planning problem of the objective function. For each objective function, calculate the corresponding satisfaction according to its value range in the Pareto frontier solution set to measure the satisfaction of each solution for different objective functions. Select the best compromise solution, compare and select the best compromise solution by maximizing the minimum satisfaction principle, and use the formula Among all Pareto frontier solutions, find the solution that maximizes the minimum satisfaction as the best compromise, thereby achieving a reasonable trade-off between multiple conflicting objectives and determining the optimal operating state of the system.

[0087] Step 4: Calculate the delay compensation strategy and robustness enhancement measures, and use two-step prediction to compensate for the calculation delay.

[0088] A two-step prediction method is used to compensate for the calculation delay, and the impact of the quad-Boost circuit on the system dynamics is fully considered during the calculation process. When predicting the system state and related variables at two future moments k+1 and k+2, the state variables of the boost circuit, the inductor current and capacitor voltage, are incorporated into the prediction model, so that the prediction results can more accurately reflect the actual operation of the system.

[0089] A two-step prediction delay compensation strategy is added, and a dynamic error correction method is used to enhance robustness, thereby comprehensively improving the system's control performance, multi-objective optimization capability, robustness and adaptability to computational delays under complex working conditions.

[0090] Predict the system state and related variables at two future moments k+1 and k+2, correct and adjust the dynamic error, consider the impact of the quad-Boost circuit on the current when calculating the error between the predicted current and the actual measured current, and correct the error calculation method to make it more accurately reflect the actual error of the system; based on the corrected error, the formula The predicted current is corrected, where σ1 is a weighting factor 0<σ1≤1. The error correction effect is optimized by adjusting the value of σ1, thereby improving the robustness of the system to changes in overall system parameters, including changes in boost circuit parameters and permanent magnet synchronous motor parameters.

[0091] Improved cost function f according to the impact of boost circuit dcom At time k+2, all possible switch state combinations (including the switch states in the inverter and boost circuit) are evaluated, and the switch state combination that minimizes the cost function is selected as the optimal control action at time k+1, effectively compensating for the impact of calculation delay on system performance.

[0092] For example, when predicting the stator current and speed changes of the motor, the impact of the inductor current and capacitor voltage changes of the boost circuit on the system is considered at the same time, so that the prediction results can more accurately reflect the actual operation of the system. According to the improved cost function of the boost circuit, all possible switch state combinations (including the switch tube states in the inverter and boost circuit) are evaluated at time t, and the switch state combination that minimizes the cost function is selected as the optimal control action at time t, effectively compensating for the impact of calculation delay on system performance.

[0093] Dynamic error correction adjustment: When calculating the error between the predicted current and the actual measured current, the effect of the quad-Boost circuit on the current is considered and the error calculation method is corrected to more accurately reflect the actual error of the system.

[0094] For example, due to the presence of the boost circuit, there may be a deviation between the measured value of the motor stator current and the actual value. By analyzing the working principle of the boost circuit and its impact on the current, the error calculation can be corrected to more accurately reflect the actual error situation of the system.

[0095] Based on the corrected error, the predicted current is corrected, and the correction formula is: Among them, σ1 is the weighting factor 0<σ1≤1. By adjusting the value of σ1, the error correction effect is optimized, and the system is improved to be robust to the overall system parameter changes including the boost circuit parameter changes (such as the inductance and capacitance parameter changes) and the permanent magnet synchronous motor parameter changes (such as the resistance, inductance, and permanent magnet flux changes), reduce the interference of parameter changes on the system control performance, and improve the reliability and stability of the system:

[0096] Step 5: Read the reference current and extrapolate using the Lagrangian method and Measure the relevant system variables at the current time k, including the stator current and speed of the motor, the inductor current and capacitor voltage of the boost circuit, etc. These measured values ​​will serve as the basic data for subsequent calculation and control.

[0097] Calculate voltage and state prediction, for 27 possible inverter switch state combinations and different switch state combinations of power switch tubes in the boost circuit, calculate the output voltage of the quadruple boost circuit and the inverter output voltage under αβ coordinates

[0098] The modified discrete state space model is used to predict the system state at time k+1 and k+2, the neutral point voltage and Calculate the switch state change Get the common mode voltage u from the lookup table cmv (i) In the calculation process, the previously established unified dynamic model and related formulas are fully utilized and combined with the measured data for accurate calculation.

[0099] Select the optimal control action, use the compromise planning method, fuzzy decision-making and minimum and maximum satisfaction principle method to select the cost function f dcom The minimum switch state combination (including the switch states of the inverter and the boost circuit) is used as the optimal control action and is applied to the inverter and the quad-Boost circuit respectively to achieve precise control of the entire system.

[0100] By repeating the above steps, the optimal control action is selected according to the real-time status of the system in each control cycle, so that the system can operate stably and efficiently under complex working conditions, meet the optimization control requirements of the permanent magnet synchronous motor, and take into account the performance optimization of the boost circuit and the robustness of the system.

[0101] The above embodiments are only illustrative of the principles and effects of the present invention, as well as some embodiments of its application. For those skilled in the art, several modifications and improvements may be made without departing from the creative concept of the present invention, and all of these belong to the protection scope of the present invention.

Claims

1. A multi-objective robust model predictive control method for a permanent magnet synchronous motor, used in a three-level inverter-powered permanent magnet synchronous motor system with a quadruple BOOST boost circuit, characterized in that: The following steps are included S1, establish a system dynamic model including a quad-Boost circuit and a permanent magnet synchronous motor, use the forward Euler method to discretize the entire system dynamic model, obtain a discrete state space model, and construct a multi-objective cost function including current tracking, neutral point voltage, common mode voltage, switching frequency, quad-Boost circuit output voltage stability and efficiency; S2, based on the current tracking control, neutral point voltage control, common mode voltage control and switching frequency control of the inverter, the output voltage stability and efficiency of the quadruple boost circuit are considered, and the multi-objective optimization problem is solved based on compromise planning and fuzzy decision-making methods; S3, firstly, the solution set on the Pareto front of the multi-objective problem is solved by measuring the difference between the optimal objective function value and the benchmark value, then the satisfaction of the solution on the Pareto front is evaluated by fuzzy decision analysis, and finally the best compromise solution is compared and selected by maximizing the minimum satisfaction principle; S4, predict the system state and related variables at two future moments k+1 and k+2, correct and adjust the dynamic error, and based on the corrected error, use the formula Correct the predicted current, where σ1 is a weighting factor 0<σ1≤1, and the error correction effect is optimized by adjusting the value of σ1; S5, read reference current and extrapolate using the Lagrangian method and Measure the relevant system variables at the current time k, calculate the output voltage of the quadruple boost circuit and the inverter output voltage under the αβ coordinates The modified discrete state space model is used to predict the system state at time k+1 and k+2, based on the neutral point voltage and Calculate the switch state change Select f dcom The minimum switch state combination is used as the optimal control action and is applied to the inverter and quad-Boost circuit respectively to achieve precise control of the entire system.

2. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 1, characterized in that: In the step S1, a differential dynamic model of the permanent magnet synchronous motor in the dq rotating reference frame is established: in is the stator dq voltage, i dq is the stator dq current, R s is the stator winding resistance, L d and L q is the stator inductance dq, ω e is the electrical rotation frequency, Ψ pm is the permanent magnetic flux, T e and T L are electromagnetic torque and load torque, Z p is the number of extreme pairs, J m is the moment of inertia, B v is the friction coefficient, T s is the sampling interval; Electromagnetic Torque of Permanent Magnet Synchronous Motor Where L d =L q =L s ; The inductor current state equation of the quadruple Boost circuit in continuous current mode is established as: The capacitor voltage state equation is: The neutral point voltage dynamic equation is obtained Where C = C dc1 =C dc2 is the support capacitor.

3. A multi-objective robust model predictive control method for a permanent magnet synchronous motor according to claim 2, characterized in that: In step S1, the forward Euler method is used to discretize the entire system dynamic model to obtain a discrete state space model for subsequent predictive control calculation: m (k+1)=A d (k)x m (k)+B d u m (k)+H d (k), where A d (k), Bd and H d (k) are the system matrix, input matrix and feedback matrix respectively: Define the neutral point voltage prediction equation:

4. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 3, characterized in that: The current tracking control in step S2 is to calculate the predicted reference current by the Lagrangian method, and construct the cost function term according to the error between the reference current and the reference current. Calculate the neutral point voltage control cost function term based on the predicted value and actual value of the neutral point voltage The common mode voltage control cost function term is constructed based on the difference between the predicted value of the common mode voltage and the expected limit value. By calculating the change in switch state And weighted, get the switching frequency control cost function term 5. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 4, characterized in that: For the quadruple boost circuit output voltage stability, add a cost function term related to the deviation between the predicted value of the boost circuit output voltage and the set reference value Considering the efficiency of the quad-boost circuit at the same time, the cost function related to efficiency is constructed by analyzing the power loss in the circuit.

6. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 5, characterized in that: The constraints considered in the multi-objective compromise planning process in step S3 are: in is the Chebyshev distance, is the minimum value of the objective function, ω is the weight coefficient, n and N are the current number of iterations and the maximum number of iterations respectively.

7. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 6, characterized in that: The following formula is used to further search for the best compromise: Where S k n is the satisfaction of the objective function for the nth planning problem, f k,min and f k,max is the minimum and maximum value of the Pareto solution of the objective function; then through the formula Compare these satisfaction levels to identify the best compromise and determine the optimal operating state of the equipment, where S n is the minimum value of the satisfaction of the nth planning decision of the objective function, S max is the maximum value among all minimum satisfaction values.

8. A method for multi-objective robust model predictive control of a permanent magnet synchronous motor according to claim 7, characterized in that: Based on the corrected error, the corrected formula The predicted current is corrected, where σ1 is the weighting factor 0<σ1≤1,

Images