Switched reluctance motor high-performance speed regulation control parameter optimization self-adjustment method under variable speed and variable load working conditions

Abstract

The invention discloses a switched reluctance motor high-performance speed regulation control parameter optimization self-adjustment method under variable speed and variable load conditions. The method comprises the following steps: constructing a mathematical model between an optimization object and an optimization target; taking n groups of load torque, rotating speed and acceleration data; a reinforcement learning algorithm is adopted to obtain corresponding optimal control parameters; a numerical fitting method is adopted to obtain function relational expressions between the optimal control parameters and the load torque, the rotating speed and the acceleration; obtaining an optimal control parameter under the current working condition; and the switched reluctance motor is controlled according to the obtained optimal control parameters, and the whole-process optimal control of the switched reluctance motor based on the variable-speed and variable-load working condition is realized. The dynamic and steady-state performance of the switched reluctance motor based on the variable-speed and variable-load working condition is effectively improved, and the method has the advantages of being simple in algorithm, easy to implement and the like.

Description

Technical Field

[0001] The present invention relates to the field of switched reluctance motor speed control, and in particular to a method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and variable load conditions. Background Art

[0002] The switched reluctance motor (SRM) is a new type of motor with a range of advantages, including a simple and robust structure, low starting current, high starting torque, and high efficiency. It has been widely used in a variety of fields, including electric vehicles and mining locomotives. Achieving high-performance speed control of SRMs requires not only the development of effective speed control methods but also the real-time determination of optimal control parameters based on changing load conditions. This is the only way to achieve optimal technical performance in the SRM speed control system.

[0003] Extensive research has been conducted on speed control methods and control parameter optimization for switched reluctance motors (SRMs). To address the difficulty of achieving optimal control using fixed control parameters in SRMs operating at variable speeds and accelerations, patent application CN119519522A, titled "High-Performance Speed ​​Control Parameter Optimization Method for SRMs Based on a Superhelical Algorithm," proposes an optimization method that adjusts optimal control parameters in real time based on changes in motor speed and acceleration, achieving high-performance full-process speed control of SRMs in variable speed and acceleration modes. However, in certain applications, such as active heave compensation control systems for marine winches, the load operates not only in variable speed and acceleration modes but also in conditions where the load changes in real time. This real-time load change directly impacts the motor's speed regulation performance, making the control parameter optimization methods proposed for stable loads inadequate. Therefore, achieving full-process optimal control under the coupled effects of variable speed, acceleration, and load is an urgent issue. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a method for optimizing and self-adjusting the speed control parameters of a switched reluctance motor under variable speed and variable load conditions, which has a simple algorithm and is easy to implement.

[0005] The technical solution of the present invention to solve the above technical problems is: a method for optimizing and self-adjusting the high-performance speed control parameters of a switched reluctance motor under variable speed and variable load conditions, comprising the following steps:

[0006] S1: Taking the control parameters of the switched reluctance motor speed control system based on the super-helical control algorithm as the optimization object, and the operating efficiency, torque ripple coefficient and steady-state control accuracy of the switched reluctance motor as the optimization targets, a mathematical model between the optimization object and the optimization target is constructed;

[0007] S2: Determine the variation range of load torque, speed and acceleration according to the technical indicators and operating conditions of the switched reluctance motor, and randomly select n sets of load torque, speed and acceleration data at equal intervals within the variation range;

[0008] S3: For each set of selected load torque, speed and acceleration data, based on the established mathematical model between the optimization object and the optimization target, a reinforcement learning algorithm is used to obtain the optimal control parameters corresponding to each set of load torque, speed and acceleration;

[0009] S4: Based on the obtained n groups of optimal control parameters and the corresponding load torque, speed, and acceleration data, a numerical fitting method is used to obtain the functional relationship between each optimal control parameter and the load torque, speed, and acceleration;

[0010] S5: Real-time acquisition of the load torque, speed, and acceleration of the switched reluctance motor during actual operation, and the optimal control parameters under the current operating conditions are obtained based on the obtained functional relationship;

[0011] S6: Control the switched reluctance motor according to the obtained optimal control parameters to achieve optimal control of the switched reluctance motor over the entire process under variable speed and variable load conditions.

[0012] In the above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, in step S1, the basic principle of the switched reluctance motor speed control system based on the super-helical control algorithm is as follows: the switched reluctance motor speed control system includes a speed control outer loop and a torque control inner loop, wherein the speed control outer loop is controlled by adjusting the actual speed of the switched reluctance motor oh With the given speed of the switched reluctance motor Compare and get the speed deviation Δ of the switched reluctance motor oh , Δ oh After being processed by the super-helical control algorithm, the reference torque of the torque control inner loop is obtained , specifically:

[0013]

[0014] Where, T L is the load torque of the switched reluctance motor, J and B are the moment of inertia of the switched reluctance motor and the friction coefficient of the switched reluctance motor, Δ oh is the speed deviation of the switched reluctance motor, Δ oh' is Δ oh The first derivative of oh' for oh The derivative of t For time, () ′Indicates the derivation of the function in the brackets, tanh() is the hyperbolic tangent function, c 、 d 、 α 、 β They are respectively the sliding surface coefficient of the superhelical control algorithm, the hyperbolic tangent function curvature gain of the superhelical control algorithm, the power term approach rate gain of the superhelical control algorithm, and the integral term approach rate of the superhelical control algorithm; α 、 β The value of satisfies the following constraints:

[0015]

[0016] The inner loop of torque control is based on the speed control outer loop. ,Will The actual torque of the switched reluctance motor Compare the torque deviation , After being processed by the direct torque control algorithm, a control signal corresponding to the power switch in the power conversion circuit is obtained. The power switch is controlled according to the obtained control signal to achieve the tracking of the actual speed of the switched reluctance motor to the time-varying given speed.

[0017] The above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, in step S1, the optimization object includes the sliding surface coefficient of the super-helical control algorithm c , the power term approach rate gain of the superhelical control algorithm α and hyperbolic tangent function curvature gain of supercoil control algorithm d ,The optimization objectives include the operating efficiency, torque ripple coefficient and steady-state control accuracy of the ,switched reluctance motor;

[0018] Operating efficiency of switched reluctance motors or The calculation formula is:

[0019]

[0020] Where, U k 、 i k 、 oh 、 m and t r Respectively represent the phase voltage, phase current, actual speed, number of phases and rotor pole pitch of the switched reluctance motor;

[0021] Torque ripple coefficient of switched reluctance motor e The calculation formula is:

[0022]

[0023] Where, T max and T min Represent the maximum torque and minimum torque of the switched reluctance motor respectively, T ave is the average torque of the switched reluctance motor;

[0024] Steady-state control accuracy of switched reluctance motor The calculation formula is:

[0025]

[0026] Where, is the given speed of the switched reluctance motor, Δ oh i For the i The deviation between the actual speed of the switched reluctance motor and the given speed collected;

[0027] Therefore, the mathematical model between the optimization object and the optimization goal is:

[0028]

[0029] Constructing a multi-objective optimization fitness function F , specifically:

[0030]

[0031] Where, k 1. k 2. k 3 are the weight coefficients of operating efficiency, torque ripple coefficient, and steady-state control accuracy, respectively, and k 1+ k 2+ k 3=1.

[0032] In the above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions, in step S3, the specific steps of using the reinforcement learning algorithm to obtain the optimal control parameters corresponding to each set of load torque, speed, and acceleration are as follows:

[0033] Set relevant parameters, including the number of optimization objects and the maximum number of iterations T , population size I , optimize the value range of the object, UB j 、 LB j Respectively represent j The upper and lower limits of the value range of the optimization object, j=1,2,3 correspond to the optimization objects respectively c , α , d ;

[0034] S 32 : Determine the learning status and learning priority of first-generation learners;

[0035] S 33 :Calculate and compare the fitness values ​​of each individual of the initial generation learner according to the multi-objective optimization fitness function, retain the minimum fitness value and the corresponding learning state, and record the retained learning state as the best state. x best [ j ] corresponds to j The best state of the learners of each optimization object is obtained, and the learners of each optimization object are arranged in descending order according to the size of the learning priority;

[0036] S 34 : Update the learning status and learning priorities of the next generation of learners;

[0037] S 35 : Arrange the learning priorities corresponding to each learner in descending order;

[0038] S 36 : Comparison step S 34 Is the minimum fitness value of the learner individual less than the minimum fitness value retained? If so, update the minimum fitness value retained and the corresponding learner state; otherwise, the minimum fitness value retained and the corresponding learner state remain unchanged.

[0039] S 37 : Determine whether the number of iterations has reached the maximum number of iterations T If it is not reached, the number of iterations is increased by 1 and the process returns to step S. 34 ; Otherwise, output the optimal learner state and calculate the optimal value of each optimization object.

[0040] The above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, the step S 32 In , the calculation formula of the learning status and learning priority of the first-generation learners is:

[0041]

[0042] Where x i，0 [j] For the first generation j The first optimization object i The learning status of a learner, i =1,2…… I, schedule i，0 [ j ] is the first generation j The first optimization object i The learning priorities of each learner, rand () is a random number between 0 and 1.

[0043] The above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, the step S 34 In , the formula for updating the learning status and learning priority of the next generation learner is:

[0044]

[0045] Where, x i,d [ j ] indicates the d Daidi j Optimization object i The learning status of a learner, Indicates the sorted d -1st generation j Optimization object i The learning priorities of each learner, schedule i,d [ j ] indicates the d Daidi j Optimization object i The learning priorities of each learner, Indicates the d Daidi j Optimization object i The fitness function value corresponding to each learner is, l Adaptively select thresholds for individual learning priorities.

[0046] The above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, the step S 34 middle, l The expression is:

[0047]

[0048] The above-mentioned method for optimizing and self-adjusting the high-performance speed control parameters of the switched reluctance motor under variable speed and load conditions, the step S 37 In , the optimal value of each optimization object is:

[0049] .

[0050] The step S4, the numerical fitting method is adopted to obtain the function relation formula between each optimal control parameter and load torque, speed and acceleration, and the specific steps are as follows:

[0051]

[0052] In the formula, A 0, A 1, A 2, A 3, A 4, B 0, B 1, B 2, B 3, B 4, C 0, C 1, C 2, C 3, C 4, C 5, C 6, C 7, C 8 are constant coefficients.

[0053] The beneficial effects of the present application are:

[0054] 1. The present application is based on the super-coil control algorithm of the switched reluctance motor speed regulation system, and the reinforcement learning algorithm is used to optimize the control parameters to obtain the corresponding optimal control parameters, so that the speed regulation system can obtain the best running efficiency, torque ripple coefficient and steady control accuracy under any load torque, speed and acceleration.

[0055] 2. The present application determines the function relation formula between each optimal control parameter of the switched reluctance motor speed regulation system based on the super-coil control algorithm and the motor speed, acceleration and load torque, and according to each function relation formula and real-time acquisition of the speed, acceleration and load torque of the switched reluctance motor during actual operation, the optimal control parameters under the current working condition can be obtained, so as to realize the whole process optimal control of the speed regulation system under the complex working condition of variable speed, variable acceleration and variable load multi-factor coupling. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 The overall flowchart of the present application.

[0057] Figure 2 The principle block diagram of the switched reluctance motor speed regulation system based on the super-coil control algorithm of the present application. DETAILED DESCRIPTION

[0058] The application will be further described below in conjunction with the drawings and examples.

[0059] As shown in the drawings, Figure 1 a high-performance speed regulation control parameter optimization self-adjustment method for a switched reluctance motor under variable speed and variable load conditions, comprising the following steps:

[0060] S1: Taking the control parameters of the switched reluctance motor speed regulation system based on the super-helix control algorithm as the optimization object, and taking the operating efficiency, torque ripple coefficient and steady-state control accuracy of the switched reluctance motor as the optimization target, a mathematical model between the optimization object and the optimization target is constructed.

[0061] As shown in the drawings, Figure 2 the basic principle of the switched reluctance motor speed regulation control system based on the super-helix control algorithm is that the switched reluctance motor speed regulation control system includes a speed control outer loop and a torque control inner loop, wherein the speed control outer loop compares the actual speed of the switched reluctance motor oh with the given speed of the switched reluctance motor to obtain the speed deviation of the switched reluctance motor Give , Give After the super-helix control algorithm is processed, the reference torque of the torque control inner loop is obtained, which is specifically:

[0062]

[0063] In the formula, T L is the load torque of the switched reluctance motor, J and B are the moment of inertia of the switched reluctance motor and the friction coefficient of the switched reluctance motor, respectively, oh is the speed deviation of the switched reluctance motor, oh' is the first derivative of oh , oh' is the derivative of oh , t is time, and ′ represents the derivative of the function in the parentheses, tanh() is the hyperbolic tangent function, c , d , α , β are the sliding mode surface coefficient of the super-helix control algorithm, the hyperbolic tangent function curvature gain of the super-helix control algorithm, the power term approach rate gain of the super-helix control algorithm, and the integral term approach rate of the super-helix control algorithm, respectively; wherein α , β the values of satisfy the following constraint conditions:

[0064]

[0065] The torque control inner loop is based on the speed control outer loop to obtain , the torque error is obtained by comparing the actual torque of the switched reluctance motor with the given torque , , The control signals of the power switches in the power conversion circuit are obtained by the direct torque control algorithm, and the power switches are controlled according to the obtained control signals to realize the tracking of the actual speed of the switched reluctance motor to the time-varying given speed.

[0066] The optimization objects include the sliding mode surface coefficient of the super-spiral control algorithm c , the power term approaching rate gain of the super-spiral control algorithm α , and the hyperbolic tangent function curvature gain of the super-spiral control algorithm d The optimization objects include the operation efficiency of the switched reluctance motor, the torque ripple coefficient of the switched reluctance motor, and the steady-state control accuracy of the switched reluctance motor

[0067] The calculation formula of the operation efficiency of the switched reluctance motor is as follows: or

[0068]

[0069] In the formula, V, I, ω, p and θ respectively represent the phase voltage, the phase current, the actual speed, the phase number and the rotor pole pitch of the switched reluctance motor U k , i k , oh , m and t r respectively represent the phase voltage, the phase current, the actual speed, the phase number and the rotor pole pitch of the switched reluctance motor

[0070] The calculation formula of the torque ripple coefficient of the switched reluctance motor is as follows: e

[0071]

[0072] In the formula, T max and T min respectively represent the maximum torque and the minimum torque of the switched reluctance motor T max and T min respectively represent the maximum torque and the minimum torque of the switched reluctance motor T ave is the average torque of the switched reluctance motor

[0073] The calculation formula of the steady-state control accuracy of the switched reluctance motor is as follows:

[0074]

[0075] In the formula, ω ref is the given speed of the switched reluctance motor, and Δω is the torque error ​​​​ oh i the deviation between the actual rotation speed of the switch reluctance motor collected for the first time and the given rotation speed; i the deviation between the actual rotation speed of the switch reluctance motor collected for the first time and the given rotation speed;

[0076] Therefore, the mathematical model between the optimization object and the optimization target is:

[0077]

[0078] constructing a multi-objective optimization fitness function F , specifically:

[0079]

[0080] In the formula, k 1, k 2, k 3 are weight coefficients of the running efficiency, the torque ripple coefficient and the steady-state control accuracy respectively, and k 1+ k 2 + k 3=1.

[0081] S2: determining the variation ranges of the load torque, the rotation speed and the acceleration according to the technical indexes and the operation conditions of the switch reluctance motor, and taking a plurality of groups of load torque, rotation speed and acceleration data at equal intervals within the variation ranges. n

[0082] S3: for each group of load torque, rotation speed and acceleration data selected, obtaining the optimal control parameters corresponding to each group of load torque, rotation speed and acceleration according to the mathematical model between the optimization object and the optimization target established.

[0083] The specific steps of obtaining the optimal control parameters corresponding to each group of load torque, rotation speed and acceleration by using the reinforcement learning algorithm are as follows:

[0084] S 31 : setting relevant parameters, including the number of optimization objects, the maximum number of iterations T , the population size I , the value range of the optimization object, UB j 、 LB j , respectively representing the upper limit value and the lower limit value of the value range of the first j optimization object, j =1, 2, 3 respectively corresponding to the optimization object c , α , d ;

[0085] S 32 ​: Determine the learning status and learning priority of the first-generation learners. The calculation formula is:

[0086]

[0087] Where x i，0 [j] For the first generation j The first optimization object i The learning status of a learner, i =1,2…… I , schedule i，0 [ j ] is the first generation j The first optimization object i The learning priorities of each learner, rand () is a random number between 0 and 1.

[0088] S 33 :Calculate and compare the fitness values ​​of each individual of the initial generation learner according to the multi-objective optimization fitness function, retain the minimum fitness value and the corresponding learning state, and record the retained learning state as the best state. x best [ j ] corresponds to j The best state of the learners of each optimization object is obtained, and the learners of each optimization object are arranged in descending order according to the size of the learning priority;

[0089] S 34 : Update the learning status and learning priority of the next generation of learners. The formula is:

[0090]

[0091] Where, x i,d [ j ] indicates the d Daidi j Optimization object i The learning status of a learner, Indicates the sorted d -1st generation j Optimization object i The learning priorities of each learner, schedule i,d [ j ] indicates the d Daidi j Optimization object i The learning priorities of each learner, Indicates the d Daidi j Optimization objecti The fitness function value corresponding to each learner, l Adaptively select thresholds for individual learning priorities:

[0092]

[0093] S 35 : Arrange the learning priorities corresponding to each learner in descending order;

[0094] S 36 : Comparison step S 34 Is the minimum fitness value of the learner individual less than the minimum fitness value retained? If so, update the minimum fitness value retained and the corresponding learner state; otherwise, the minimum fitness value retained and the corresponding learner state remain unchanged.

[0095] S 37 : Determine whether the number of iterations has reached the maximum number of iterations T If it is not reached, the number of iterations is increased by 1 and the process returns to step S. 34 ; Otherwise, output the optimal learner state and calculate the optimal value of each optimization object;

[0096] The optimal value of each optimization object is:

[0097] .

[0098] S4: Based on the results n The optimal control parameters and the corresponding load torque, speed and acceleration data are set, and the functional relationship between each optimal control parameter and the load torque, speed and acceleration is obtained by numerical fitting method.

[0099] The functional relationships between the optimal control parameters and the load torque, speed and acceleration are obtained by numerical fitting method, as follows:

[0100]

[0101] Where, A 0. A 1. A 2. A 3. A 4. B 0. B 1. B 2. B 3. B 4. C 0. C 1. C 2. C 3. C 4. C 5. C6. C 7. C 8 are all constant coefficients.

[0102] S5: The load torque, speed and acceleration of the switched reluctance motor during actual operation are collected in real time, and the optimal control parameters under the current working conditions are obtained based on the obtained functional relationship.

[0103] S6: Control the switched reluctance motor according to the obtained optimal control parameters to achieve optimal control of the switched reluctance motor over the entire process under variable speed and variable load conditions.

[0104] In order to illustrate the effect of the high-performance speed regulation control parameter optimization self-adjustment method of the switched reluctance motor under variable speed and variable load conditions provided by the present invention, a 12 / 8-pole switched reluctance motor is used as an example for verification. Its main technical parameters are shown in Table 1.

[0105]

[0106] According to the technical parameters shown in Table 1 and the operating conditions of the switched reluctance motor, if its rated load torque and maximum angular acceleration are 81 and 82, respectively, and 80 rad / s 2 , and within the range of rated load torque, rated speed and maximum angular acceleration, if any 2 , 60 rpm, 10 rad / s 2 As the starting point, and any 2 , 35 rpm, 2 rad / s 2 20 sets of load torque, speed and acceleration data are selected at equal intervals. For each set of selected data, the control parameter optimization self-adjustment method provided by the present invention is used to obtain the corresponding optimal control parameters, as shown in Table 2.

[0107]

[0108] According to the optimal control parameter values ​​and the corresponding load torque, motor speed and acceleration data shown in Table 2, the functional relationship between the optimal parameters and the load torque, motor speed and acceleration is obtained by numerical fitting method, which is specifically:

[0109]

[0110] To further illustrate the effectiveness of the control parameter optimization and self-adjustment method provided by the present invention (hereinafter referred to as the present method), a comparative analysis is conducted between the present method and a conventional fixed control parameter control method (hereinafter referred to as the conventional method). The control parameters used in the conventional method are the optimal control parameters obtained using a reinforcement learning algorithm for a switched reluctance motor under rated operating conditions, namely, rated load torque, rated speed, and maximum acceleration, as shown in Table 3.

[0111]

[0112] Take any two sets of data within the range of rated load torque, rated speed and maximum angular acceleration of the switched reluctance motor. For example, the load torque, speed and acceleration of the motor are: 25 , 900 rpm, 20.94 rad / s 2 and 40 , 400 rpm, 10.47 rad / s 2 When the method and the traditional method are used to control the switched reluctance motor speed control system, the operating efficiency, torque ripple coefficient and steady-state control accuracy are shown in Table 4 and Table 5 respectively.

[0113]

[0114]

[0115] It can be seen from Tables 4 and 5 that for any two sets of load torque, motor speed and acceleration data, compared with the traditional method, the motor efficiency obtained by this method is improved by 1.12% and 0.45% respectively, the torque ripple coefficient is reduced by 22.12% and 28.23% respectively, and the steady-state control accuracy is improved by 6.86% and 7.56% respectively. This further proves that the control effect of the switched reluctance motor speed control system is effectively improved by using this method.

Claims

1. A method for optimizing and self-adjusting the high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions, characterized in that: The following steps are involved: S1: Taking the control parameters of the switched reluctance motor speed control system based on the super-helical control algorithm as the optimization object, and the operating efficiency, torque ripple coefficient and steady-state control accuracy of the switched reluctance motor as the optimization targets, a mathematical model between the optimization object and the optimization target is constructed; S2: Determine the variation range of load torque, speed and acceleration according to the technical indicators and operating conditions of the switched reluctance motor, and randomly select n sets of load torque, speed and acceleration data at equal intervals within the variation range; S3: For each set of selected load torque, speed and acceleration data, based on the established mathematical model between the optimization object and the optimization target, a reinforcement learning algorithm is used to obtain the optimal control parameters corresponding to each set of load torque, speed and acceleration; S4: Based on the obtained n groups of optimal control parameters and the corresponding load torque, speed, and acceleration data, a numerical fitting method is used to obtain the functional relationship between each optimal control parameter and the load torque, speed, and acceleration; S5: Real-time acquisition of the load torque, speed, and acceleration of the switched reluctance motor during actual operation, and the optimal control parameters under the current operating conditions are obtained based on the obtained functional relationship; S6: Control the switched reluctance motor according to the obtained optimal control parameters to achieve optimal control of the switched reluctance motor over the entire process under variable speed and variable load conditions.

2. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 1 is characterized in that: In step S1, the basic principle of the switched reluctance motor speed control system based on the super-helical control algorithm is as follows: the switched reluctance motor speed control system includes a speed control outer loop and a torque control inner loop, wherein the speed control outer loop is controlled by the actual speed of the switched reluctance motor. ω With the given speed of the switched reluctance motor Compare and get the speed deviation Δ of the switched reluctance motor ω , Δ ω After being processed by the super-helical control algorithm, the reference torque of the torque control inner loop is obtained , specifically: ； Where, T L is the load torque of the switched reluctance motor, J and B are the moment of inertia of the switched reluctance motor and the friction coefficient of the switched reluctance motor, Δ ω is the speed deviation of the switched reluctance motor, Δ ω′ is Δ ω The first derivative of ω′ for ω The derivative of t For time, () ′ Indicates the derivation of the function in the brackets, tanh() is the hyperbolic tangent function, c 、 δ 、 α 、 β They are respectively the sliding surface coefficient of the superhelical control algorithm, the hyperbolic tangent function curvature gain of the superhelical control algorithm, the power term approach rate gain of the superhelical control algorithm, and the integral term approach rate of the superhelical control algorithm; α 、 β The value of satisfies the following constraints: ； The inner loop of torque control is based on the speed control outer loop. ,Will The actual torque of the switched reluctance motor Compare the torque deviation , After being processed by the direct torque control algorithm, a control signal corresponding to the power switch in the power conversion circuit is obtained. The power switch is controlled according to the obtained control signal to achieve the tracking of the actual speed of the switched reluctance motor to the time-varying given speed.

3. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 2 is characterized in that: In step S1, the optimization objects include the sliding surface coefficient c of the superhelical control algorithm, the power term approach rate gain of the superhelical control algorithm, and the α and hyperbolic tangent function curvature gain of supercoil control algorithm δ ,The optimization objectives include the operating efficiency, torque ripple coefficient and steady-state control accuracy of the ,switched reluctance motor; Operating efficiency of switched reluctance motors η The calculation formula is: ； Where, U k 、 i k 、 ω 、 m and τ r Respectively represent the phase voltage, phase current, actual speed, number of phases and rotor pole pitch of the switched reluctance motor; The calculation formula of the torque ripple coefficient ε of the switched reluctance motor is: ； Where, T max and T min Represent the maximum torque and minimum torque of the switched reluctance motor respectively, T ave is the average torque of the switched reluctance motor; Steady-state control accuracy of switched reluctance motor The calculation formula is: ； Where, is the given speed of the switched reluctance motor, Δ ω i For the i The deviation between the actual speed of the switched reluctance motor and the given speed collected; Therefore, the mathematical model between the optimization object and the optimization goal is: ； Constructing a multi-objective optimization fitness function F , specifically: ； Where, k 1. k 2. k 3 are the weight coefficients of operating efficiency, torque ripple coefficient, and steady-state control accuracy, respectively, and k 1+ k 2+ k 3=1.

4. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 3 is characterized in that: In step S3, the specific steps of using the reinforcement learning algorithm to obtain the optimal control parameters corresponding to each set of load torque, speed and acceleration are as follows: S 31 : Set relevant parameters, including the number of optimization objects and the maximum number of iterations T , population size I , optimize the value range of the object, UB j 、 LB j Respectively represent j The upper and lower limits of the value range of the optimization object, j =1,2,3 correspond to the optimization objects respectively c , α , δ ; S 32 : Determine the learning status and learning priority of first-generation learners; S 33 :Calculate and compare the fitness values ​​of each individual of the initial generation learner according to the multi-objective optimization fitness function, retain the minimum fitness value and the corresponding learning state, and record the retained learning state as the best state. x best [ j ] corresponds to j The best state of the learners of each optimization object is obtained, and the learners of each optimization object are arranged in descending order according to the size of the learning priority; S 34 : Update the learning status and learning priorities of the next generation of learners; S 35 : Arrange the learning priorities corresponding to each learner in descending order; S 36 : Comparison step S 34 Is the minimum fitness value of the learner individual less than the minimum fitness value retained? If so, update the minimum fitness value retained and the corresponding learner state; otherwise, the minimum fitness value retained and the corresponding learner state remain unchanged. S 37 : Determine whether the number of iterations has reached the maximum number of iterations T If it is not reached, the number of iterations is increased by 1 and the process returns to step S. 34 ; Otherwise, output the optimal learner state and calculate the optimal value of each optimization object.

5. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 4 is characterized in that: The step S 32 In , the calculation formula of the learning status and learning priority of the first-generation learners is: ； Where x i，0 [j] For the first generation j The first optimization object i The learning status of a learner, i =1,2…… I , schedule i，0 [ j ] is the first generation j The first optimization object i The learning priorities of each learner, rand () is a random number between 0 and 1.

6. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 5 is characterized in that: The step S 34 In , the formula for updating the learning status and learning priority of the next generation learner is: ； Where, x i,d [ j ] indicates the d Daidi j Optimization object i The learning status of a learner, Indicates the sorted d -1st generation j Optimization object i The learning priorities of each learner, schedule i,d [ j ] indicates the d Daidi j Optimization object i The learning priorities of each learner, Indicates the d Daidi j Optimization object i The fitness function value corresponding to each learner is, λ Adaptively select thresholds for individual learning priorities.

7. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 6 is characterized in that: The step S 34 middle, λ The expression is: 。 8. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 7 is characterized in that: The step S 37 In , the optimal value of each optimization object is: 。 9. The method for optimizing and self-adjusting high-performance speed control parameters of a switched reluctance motor under variable speed and load conditions according to claim 8 is characterized in that: In step S4, a numerical fitting method is used to obtain the functional relationship between each optimal control parameter and the load torque, speed and acceleration, as follows: ； In the formula, A0, A1, A2, A3, A4, B0, B1, B2, B3, B4, C0, C1, C2, C3, C4, C5, C6, C7, and C8 are all constant coefficients.

Images