PID parameter estimation method based on MIWOA

By introducing new shrinkage factors, adaptive weights, sparrow search algorithm and Cauchygaussian mutation strategy in MIWOA, the problem of cumbersome adjustment of PID controller parameters and easy to fall into local optimality is solved, and the performance and accuracy of the PID control system are improved.

CN119024677APending Publication Date: 2024-11-26XIDIAN UNIV
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411123202.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

In the prior art, the parameter adjustment of PID controllers is complicated and has poor results. The intelligent optimization algorithm has problems such as low search efficiency and easy to fall into local optimality in complex optimization problems.

Method used

Using the MIWOA-based PID parameter estimation method, the global search ability and local optimization ability are enhanced through new shrinkage factors, adaptive weights, sparrow search algorithm and Cauchygaussian mutation strategy, avoid falling into local optimization, improve convergence speed and optimization accuracy .

Benefits of technology

It significantly improves the overall performance of the PID control system, balances response speed and steady-state error, and improves the efficiency and accuracy of PID parameter optimization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024677A_ABST
    Figure CN119024677A_ABST
Patent Text Reader

Abstract

The invention relates to a PID (Proportion Integration Differentiation) parameter estimation method based on MIWOA. The method comprises the following steps: initializing the current iteration times of the MIWOA; when the current number of iterations does not reach the maximum number of iterations, obtaining a new shrinkage factor and an adaptive weight according to the current number of iterations and the maximum number of iterations; obtaining an updated position updating model according to the new shrinkage factor and the adaptive weight; according to the updated position updating model, obtaining the updated position of each individual and the corresponding fitness; and processing each individual through a sparrow search algorithm inserter mechanism or a Cauchy Gaussian mutation strategy to obtain an optimal individual, and obtaining a globally optimal solution. Through the new shrinkage factor, the adaptive weight, the sparrow search algorithm inserter mechanism and the Cauchy Gaussian mutation strategy, the global search capability and the local optimization capability of the whale optimization algorithm are enhanced, local optimum is avoided, the convergence speed and the optimization precision are significantly improved, and the overall performance of the PID control system is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of control system optimization, and in particular to a PID parameter estimation method based on MIWOA. Background Art

[0002] PID controller (Proportion Integration Differentiation) is widely used in various industrial control systems due to its simple structure, few adjustment parameters and high stability. However, in practical applications, the adjustment of PID controller parameters (i.e. Kp, Ki and Kd) directly affects the control performance of the system. Traditional PID parameter adjustment methods, such as empirical method, Ziegler-Nichols method and trial and error method, often require a lot of experience and experiments, the adjustment process is cumbersome and the effect is not necessarily optimal. Therefore, seeking an efficient and accurate PID parameter optimization method has become a research hotspot.

[0003] In recent years, intelligent optimization algorithms have been gradually applied to PID parameter optimization, such as genetic algorithm (GA), particle swarm optimization (PSO), ant colony optimization (ACO), grey wolf optimization (GWO), etc. These algorithms can search for the optimal solution in a large range by simulating the evolution or group behavior of nature, and improve the efficiency and accuracy of parameter adjustment. However, these algorithms also have some disadvantages, such as easy to fall into local optimal, slow convergence speed, and complex parameter setting.

[0004] Whale Optimization Algorithm (WOA) is an emerging intelligent optimization algorithm. WOA simulates the behavior of humpback whales encircling prey to achieve global exploration and local development of the search space. WOA has the advantages of few parameters, easy implementation, and strong global search capability, and performs well in solving continuous optimization problems. However, when dealing with complex optimization problems, standard WOA still has problems such as the search efficiency needs to be improved and it is easy to fall into local optimality. Summary of the invention

[0005] In order to solve the above problems existing in the prior art, the present invention provides a PID parameter estimation method based on MIWOA.

[0006] According to a first aspect of an embodiment of the present invention, a PID parameter estimation method based on MIWOA is provided, the method comprising:

[0007] Initialize the algorithm parameters of MIWOA; wherein the algorithm parameters include the current number of iterations;

[0008] When the current number of iterations does not reach the maximum number of iterations, obtaining a new shrinkage factor and an adaptive weight according to the current number of iterations and the maximum number of iterations;

[0009] Obtaining an updated position update model according to the new shrinkage factor and the adaptive weight; wherein the position update model is obtained according to MIWOA;

[0010] updating the position and corresponding fitness of each individual according to the updated position update model to obtain the updated position and corresponding fitness of each individual; wherein the individual includes a set of preset PID controller parameters;

[0011] According to the updated positions and corresponding fitness of each individual, each individual is processed by a joiner mechanism of a sparrow search algorithm or a Cauchy-Gauss mutation strategy to obtain the best individual;

[0012] A global optimal solution is obtained according to the position and fitness of the best individual; wherein the global optimal solution is the optimal PID controller parameters.

[0013] Optionally, the new shrinkage factor is expressed as follows:

[0014]

[0015] Wherein, a represents the new shrinkage factor, ε is the first correlation coefficient, η is the second correlation coefficient, k represents the current iteration number, k max Represents the maximum number of iterations.

[0016] Optionally, the adaptive weight is expressed as follows:

[0017]

[0018] Wherein, ω represents the adaptive weight, e is a natural constant, α is used to adjust the change of the adaptive weight vertically, and β is used to adjust the change of the adaptive weight horizontally.

[0019] Optionally, the position update model includes three search mechanisms: surrounding prey, bubble net attack and searching prey.

[0020] Optionally, updating the position of each individual and the corresponding fitness according to the updated position updating model to obtain the updated position of each individual and the corresponding fitness includes:

[0021] Get the first random number;

[0022] When the first random number is greater than or equal to a preset threshold, the updated positions and corresponding fitness of each individual are obtained according to the bubble network attack;

[0023] Get the adjustment coefficient;

[0024] When the first random number is less than the preset threshold value and the adjustment coefficient is greater than or equal to the adjustment coefficient, the updated position and corresponding fitness of each individual are obtained according to the search for prey;

[0025] When the first random number is smaller than the preset threshold value, and the adjustment coefficient is smaller than the adjustment coefficient, the updated positions of each individual and the corresponding fitness are obtained according to the surrounded prey.

[0026] Optionally, the encirclement of the prey refers to the following formula:

[0027] X(k+1)=ω·X*(k)-A·D1;

[0028] Where X(k+1) represents the position and corresponding fitness of each individual after the k+1th iteration, k represents the number of iterations, ω is the adaptive weight, and X * (k) represents the current best individual position vector, D1 = |C·X * (k)-X(k)| represents the distance between other individuals and the current best individual, || represents the absolute value, A is the first coefficient vector, C is the second coefficient vector, and X(k) represents the position of each individual after the kth iteration and the corresponding fitness.

[0029] Optionally, the bubble network attack refers to the following formula:

[0030] X(k+1)=ω·X * (k)+D2·e bl ·cos(2πl);

[0031] Where D2 = |X * (k)-X(k)| represents the distance between other individuals and the current best individual, b is a preset constant, l is a random number between [-1,1], and e is a natural constant.

[0032] Optionally, the search for prey refers to the following formula:

[0033] X(k+1)=ω·X rand (k)-A·D3;

[0034] Among them, X rand (k) represents the position vector of a random individual randomly selected from the current individuals, D3 = |C·X rand (k)-X(k)| represents the distance between other individuals and the random individual.

[0035] Optionally, the step of processing each individual according to the updated position and corresponding fitness of each individual by using a joiner mechanism of a sparrow search algorithm or a Cauchy-Gauss mutation strategy to obtain the best individual includes:

[0036] Get a second random number;

[0037] When the second random number is less than the disturbance probability, each of the individuals is processed by the joiner mechanism of the sparrow search algorithm;

[0038] When the second random number is greater than or equal to the disturbance probability, each individual is processed by the Cauchy-Gaussian mutation strategy.

[0039] Optionally, when the second random number is less than the disturbance probability, processing each individual through the joiner mechanism of the sparrow search algorithm includes:

[0040] When the second random number is less than the disturbance probability, generating a third random number;

[0041] When the third random number is less than the safety threshold, each of the individuals is processed by the first formula of the joiner mechanism of the sparrow search algorithm;

[0042] When the third random number is greater than or equal to the security threshold, each of the individuals is processed by the second formula of the joiner mechanism of the sparrow search algorithm.

[0043] The technical solution provided by the present invention may include the following beneficial effects:

[0044] Through the above technical solution, the present invention designs a new shrinkage factor to obtain MIWOA, and introduces adaptive weights, the sparrow search algorithm joiner mechanism and the Cauchy-Gauss mutation strategy, which enhances the global search capability and local optimization capability of the whale optimization algorithm, avoids falling into the local optimum, and significantly improves the convergence speed and optimization accuracy. Through fitness, the optimization process better balances the response speed and steady-state error, and improves the overall performance of the PID control system.

[0045] Other features and advantages of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention but do not constitute a limitation of the present invention. In the accompanying drawings:

[0047] Figure 1 The figure is a flow chart of a PID parameter estimation method based on MIWOA according to an exemplary embodiment.

[0048] Figure 2 It is a flowchart of another MIWOA-based PID parameter estimation method according to an exemplary embodiment.

[0049] Figure 3 is a block diagram of a PID control structure according to an exemplary embodiment.

[0050] Figure 4 is a diagram showing an adaptive change curve of a disturbance probability according to an exemplary embodiment.

[0051] Figure 5 is a flowchart of PID parameter estimation based on MIWOA according to an exemplary embodiment. DETAILED DESCRIPTION

[0052] In order to facilitate the understanding of the scheme of the present invention, the relevant state of the prior art and the inventive concept of the present invention are first briefly described.

[0053] In order to further improve the performance of WOA, researchers have proposed a variety of improvement strategies in recent years, such as introducing chaotic search and mutation operations. These improvement strategies improve the convergence speed and optimization accuracy of the algorithm by enhancing the global search capability and local development capability of WOA. However, the existing WOA cannot be well applied to solve the problem of PID parameter selection in actual control systems. Some of its variants still find it difficult to avoid premature convergence and insufficient convergence accuracy when solving PID parameter problems, and cannot provide reliable results. Therefore, the present invention proposes a PID parameter estimation method based on MIWOA to further improve the efficiency and accuracy of PID controller parameter optimization.

[0054] Figure 1 is a flow chart of a PID parameter estimation method based on MIWOA according to an exemplary embodiment. Figure 1 As shown, the method includes the following steps.

[0055] S101. Initialize algorithm parameters of MIWOA; wherein the algorithm parameters include the current number of iterations.

[0056] It is understandable that the relevant parameters of initializing MIWOA (Multi-strategy Improved Whale Optimization Algorithm) include population size N, current number of iterations k, maximum number of iterations k max , create a i A population of individuals, where X i Represented as a random value (i=1,2,...,N).

[0057] S102: When the current number of iterations does not reach the maximum number of iterations, a new shrinkage factor and an adaptive weight are obtained according to the current number of iterations and the maximum number of iterations.

[0058] It is worth mentioning that Figure 2 is a flowchart of another PID parameter estimation method based on MIWOA according to an exemplary embodiment. Figure 2 As shown, before judging whether the current number of iterations reaches the maximum number of iterations, the population fitness can be calculated according to the algorithm parameters of the initialized MIWOA, and the initial best individual can be found. When the current number of iterations reaches the maximum number of iterations, the initial best individual is output as the best individual.

[0059] It can be understood that the contraction factor a′ of the basic whale optimization algorithm decreases linearly from 2 to 0 during the population iteration cycle, that is:

[0060]

[0061] The new shrinkage factor designed by the present invention is expressed as follows:

[0062]

[0063] Among them, a represents the new shrinkage factor, ε is the first correlation coefficient, η is the second correlation coefficient, ε and η can adjust the iteration ratio between exploration and development, k represents the current iteration number, k max Indicates the maximum number of iterations. Considering that the higher the frequency of trigonometric functions is, the higher the computational cost may be. The higher the power of the exponential function, the greater the impact on the range of curve convergence. After multiple experiments to adjust the parameters, the algorithm performs best when ε=0.5 and η=0.3.

[0064] In addition, an adaptive weight is designed and introduced in MIWOA, and the adaptive weight is expressed as follows:

[0065]

[0066] Among them, ω represents the adaptive weight, e is a natural constant, α is used to adjust the adaptive weight change vertically, and β is used to adjust the adaptive weight change horizontally. Depending on the difference in α, the improved whale optimization algorithm obtains different weights when the number of iterations is the same. These two parameters coordinately adjust the weights to ensure the flexibility of the adaptive weights.

[0067] S103. Obtain an updated position update model according to the new shrinkage factor and the adaptive weight; wherein the position update model is obtained according to MIWOA.

[0068] It can be understood that the position update model includes three search mechanisms: surrounding prey, bubble net attack and searching prey.

[0069] The following formula is used to surround the prey:

[0070] X(k+1)=ω·X*(k)-A·D1;

[0071] Among them, X(k+1) represents the position and corresponding fitness of each individual after the k+1th iteration, k represents the number of iterations, ω is the adaptive weight, and X * (k) represents the current best individual position vector, D1 = |C·X * (k)-X(k)| represents the distance between other individuals and the current best individual, || represents the absolute value, A is the first coefficient vector, C is the second coefficient vector, and X(k) represents the position of each individual after the kth iteration and the corresponding fitness.

[0072] The first coefficient vector reference is represented as follows:

[0073] A = 2a·r1-a;

[0074] The second coefficient vector reference is represented as follows:

[0075] C = 2·r2;

[0076] Where r1 and r2 both represent random numbers between [0,1].

[0077] The bubble net attack refers to the following formula:

[0078] X(k+1)=ω·X * (k)+D2·e bl ·cos(2πl);

[0079] Where D2 = |X * (k)-X(k)| represents the distance between other individuals and the current best individual, b is a preset constant, l is a random number between [-1,1], and e is a natural constant.

[0080] Searching for prey refers to the following formula:

[0081] X(k+1)=ω·X rand (k)-A·D3;

[0082] Among them, X rand (k) represents the position vector of a random individual randomly selected from the current individuals, D3 = |C·X rand (k)-X(k)| represents the distance between other individuals and the random individual.

[0083] S104, updating the position and corresponding fitness of each individual according to the updated position update model, to obtain the updated position and corresponding fitness of each individual; wherein the individual includes a set of preset PID controller parameters.

[0084] It can be understood that the present invention regards a set of PID controller parameters as an individual of MIWOA, and a plurality of individuals form a population.

[0085] In one embodiment, Figure 3 is a PID control structure block diagram according to an exemplary embodiment, and the formula of the PID controller is:

[0086]

[0087] Among them, u(t) is the output value, t represents time, K p , K i , K d are the proportional parameter, integral parameter and differential parameter of the PID controller respectively. e(t) is the error, that is, the difference between the expected value and the output. The proportional parameter K p It can improve the response speed of the system and reduce the steady-state error, but it cannot completely eliminate the steady-state error, making the system unable to reach the predetermined point. Integral parameter K i It can eliminate steady-state errors and improve accuracy. i When it is too large, the system may overshoot and oscillate, making the system response slower. d By utilizing the deviation change trend, adjustments can be made in advance to reduce system overshoot and oscillation and improve system stability.

[0088] According to the above content, the evaluation indicators are mainly steady-state error and response time, as follows:

[0089] Steady-state error mainly refers to the difference between the system output value and the set value after the system reaches a stable state. Ideally, the steady-state error is zero.

[0090] e(t)=lim t→∞ (r(t)-u(t));

[0091] Among them, r(t) is the set value and u(t) is the system output value.

[0092] Response time is the time required for the system output to reach the set value from the initial state and stabilize. It usually includes the rise time t r , peak time t p and settling time t s .t rIt is mainly the time required for the system output to go from 10% to 90% of the set value. p It is mainly the time required for the system output to reach its maximum value for the first time. s It mainly refers to the time required for the system output to stabilize within a certain error band of the set value (usually 5% or 2%).

[0093] The system error and IAE function (Integral Absolute Error) are used as fitness functions and solved by MIWOA.

[0094] The IAE function reference is shown in the following formula:

[0095]

[0096] Where e(t) is the system error.

[0097] Optionally, S104 may include:

[0098] Get the first random number;

[0099] When the first random number is greater than or equal to a preset threshold, the updated positions and corresponding fitness of each individual are obtained according to the bubble network attack;

[0100] Get the adjustment coefficient;

[0101] When the first random number is less than a preset threshold value and the adjustment coefficient is greater than or equal to the adjusted coefficient, the updated positions and corresponding fitness of each individual are obtained according to the search for prey;

[0102] When the first random number is smaller than a preset threshold value and the adjustment coefficient is smaller than the adjusted coefficient, the updated positions of each individual and the corresponding fitness are obtained according to the surrounding prey.

[0103] It is understandable that if Figure 2 As shown, when using MIWOA, it is assumed that whales will choose between shrinking prey, surrounding prey and bubble net attack behaviors with equal probability when hunting. The mathematical model is as follows:

[0104]

[0105] Wherein, p is the first random number.

[0106] Surrounding prey and searching prey are mainly determined by the adjustment coefficient A. For example, the value range of the adjustment coefficient A is [2, -2]. When A ≥ 1, the position and corresponding fitness of each individual after update are obtained by searching prey. When A < 1, the position and corresponding fitness of each individual after update are obtained by surrounding prey.

[0107] S105. According to the updated positions and corresponding fitness of each individual, each individual is processed by the joiner mechanism of the sparrow search algorithm or the Cauchy-Gauss mutation strategy to obtain the best individual.

[0108] It is understandable that when the number of iterations reaches the current maximum number of iterations, the algorithm is stopped, the entire population can be sorted according to the fitness value of each individual, and each individual can be processed through the sparrow search algorithm joiner mechanism or the Cauchy-Gauss mutation strategy to obtain the best individual.

[0109] Optionally, S105 may include:

[0110] Get a second random number;

[0111] When the second random number is less than the perturbation probability, each individual is processed through the joiner mechanism of the sparrow search algorithm;

[0112] When the second random number is greater than or equal to the perturbation probability, each individual is processed by the Cauchy-Gauss mutation strategy.

[0113] Understandably, Figure 4 is an adaptive change curve diagram of a disturbance probability according to an exemplary embodiment. Figure 4 As shown, generate a second random number r1 between [0,1], and determine whether the second random number r1 is less than the disturbance probability P V If r1 <P V , select the sparrow search algorithm to join the mechanism to search, otherwise select the Cauchy-Gauss mutation strategy to search. When performing global exploration in the early stage, it is necessary to find as many possible solutions as possible, and in the later local development, potential better solutions should be explored around the current solution. Therefore, the perturbation strategy adopted by the present invention should be to adjust the perturbation size adaptively following the algorithm iteration. The perturbation value is larger in the early stage of the algorithm, which is conducive to enhancing the global exploration, and the perturbation value is smaller in the later stage, which is conducive to the algorithm to perform local exploration near the current optimal solution. The intuitive feeling is that the perturbation value will decrease with the iteration process. The present invention adopts a cosine-based adaptive perturbation strategy to further balance the exploration and development of the algorithm.

[0114] The disturbance probability formula is as follows:

[0115]

[0116] Among them, k is the current iteration number, k max is the maximum number of iterations.

[0117] Optionally, when the second random number is less than the disturbance probability, processing each individual through the joiner mechanism of the sparrow search algorithm may include:

[0118] When the second random number is less than the perturbation probability, a third random number is generated;

[0119] When the third random number is less than the safety threshold, each individual is processed by the first formula of the joiner mechanism of the sparrow search algorithm;

[0120] When the third random number is greater than or equal to the safety threshold, each individual is processed by the second formula of the joiner mechanism of the sparrow search algorithm.

[0121] Furthermore, when the second random number is less than the perturbation probability, that is, r1 < P V At this time, the joiner mechanism of the sparrow search algorithm is used to perturb it. Define the individual with the minimum fitness as the optimal position of the joiner of the sparrow search algorithm. Then, a third random number r2 between [0, 1] is generated, and it is judged whether the third random number r2 is less than the safety threshold ST. If r2 < ST, the first formula of the joiner of the sparrow search algorithm is used for position update, otherwise the second formula of the joiner of the sparrow search algorithm is used for position update. The joiner mechanism of the sparrow search algorithm avoids the phenomenon that the whale optimization algorithm is overly clustered in the later stage of the population and falls into the extreme value. The range of the safety threshold ST is between [0.5, 1]. For example, ST = 0.7 can be taken, that is, 70% of the individuals in the population tend to the direction of the population optimal position.

[0122] The position update formula of the joiner mechanism of the sparrow search algorithm is as follows:

[0123]

[0124] Among them, Q describes a normally distributed random number, X i (k) represents the mutated position of the i-th individual at the k-th iteration, X P represents the best position occupied by the discoverer, X worst represents the global worst position of the population in the iteration. I represents a 1×d matrix, d is the dimension of the problem, and each element in the matrix is randomly assigned 1 or -1, and + = I T (II T ) -1 . L represents a matrix with all elements being 1. When i > N / 2, it means that the individual i does not obtain a favorable position and needs to expand the flight area to find food elsewhere. When i ≤ N / 2, it means that the joiner i will forage around the best position and perform local search.

[0125] Furthermore, as Figure 2 shown, when the second random number is greater than or equal to the perturbation probability, that is, r1 ≥ P V At this time, the algorithm uses the Cauchy-Gaussian mutation strategy for search to avoid the algorithm falling into local optimum.

[0126] The Cauchy-Gauss mutation strategy is as follows:

[0127]

[0128] in, is the position after mutation, Cauchy(0,1) is a random number that satisfies the standard Cauchy distribution, Gauss(0,1) is a random number that satisfies the Gaussian distribution, It is a dynamic parameter that is adaptively adjusted with the number of iterations.

[0129] S106. Obtain a global optimal solution according to the position and fitness of the best individual; wherein the global optimal solution is the optimal PID controller parameters.

[0130] It can be understood that the position of the best individual in the population and its fitness value are output to obtain the global optimal solution; that is, after the iteration, the individual position with the best value of the fitness function (IAE), that is, the optimal PID controller parameter (K p , K i , K d ).

[0131] In one embodiment, Figure 5 The PID parameter estimation flow chart based on MIWOA is shown according to an exemplary embodiment. The PID control process is a control method based on feedback. By adjusting the input of the system, the output of the system is made as close to the set value as possible. The PID controller generates a control signal based on the current error, the historical accumulated value of the error, and the change trend of the error, and then adjusts the behavior of the system. Figure 5 It shows that the MIWOA algorithm proposed by the present invention is used to optimize the parameter K in the PID controller. p , K i and K d Specifically, IAE is used as the optimization objective function of this parameter estimation, that is, the fitness function. MIWOA initializes a set of random PID parameter sets (K p , K i and K d) as a population of initial solutions, these parameter sets represent the initial positions of the whales. For each PID parameter set, use them to configure the PID controller and simulate the control system. Calculate the system response and evaluate the fitness value based on the selected performance indicator (such as IAE). MIWOA updates each parameter set by simulating the predation behavior of whales. As the iterations proceed, the whale population gradually converges to the global optimal region. At the termination of the algorithm, the PID parameter set with the best fitness value is selected as the optimal solution. The actual controller is configured using the optimal PID parameters obtained by optimization to ensure that the system output u(t) can follow the set value r(t) and minimize the error. By optimizing the PID parameters, the controller output u(t) can better adjust the actuator or control element, thereby affecting the behavior of the system. Ideally, the optimized PID parameters enable the system to respond quickly to the set value changes, reduce overshoot, reduce steady-state error, and remain stable under disturbances. This optimization process ensures that the system output u(t) tracks the set value r(t) more accurately and stably.

[0132] In one embodiment, in order to be more suitable for practical application control systems, a PID parameter estimation method based on multi-strategy MIWOA proposed in the present invention is applied to the speed regulation system of a DC motor. The speed regulation system of a DC motor is generally controlled by a PID controller, wherein the multi-strategy improved whale optimization algorithm proposed in the present invention can be used to select the best parameters to produce the best response. The parameter values ​​of the DC motor used as a case study are listed in Table 1:

[0133] Table 1

[0134]

[0135] Among them, R a is the armature resistance, L a is the armature winding inductance, J is the equivalent moment of inertia of the motor and load based on the motor shaft, D is the equivalent friction coefficient of the motor and load based on the motor shaft, K is the motor torque constant, K b Represents the back electromotive force constant. The transfer function of the DC motor closed-loop speed control system is shown below:

[0136]

[0137] Table 2 lists the optimal PID controller parameter values ​​for optimizing DC motor speed regulation using MIWOA versus standard WOA, and other related algorithms used in the comparative study. MIWOA optimizes a single objective, which is IAE. The solution is the optimal parameters of the PID controller that achieves the minimum IAE. Other specifications, such as rise time and overshoot, are measured based on the estimated parameters of MIWOA and other algorithms in the comparison.

[0138] Table 2

[0139]

[0140] As can be seen from Table 2, for overshoot measurement, MIWOA provides significantly smaller overshoot than other algorithms. For rise time, according to the results in Table 2, MIWOA produces the smallest rise time, while PSO provides the largest rise time. For fitness value, MIWOA finds the smallest result. This result shows that MIWOA outperforms WOA and other algorithms in finding the minimum IAE and reasonable overshoot and rise time. From the example parameter estimation results, MIWOA outperforms WOA and other comparative studies in estimating the effective parameters of PID control, thereby achieving efficient IAE.

[0141] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

[0142] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0143] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the present invention, they should also be regarded as the contents disclosed by the present invention.

Claims

1. A PID parameter estimation method based on MIWOA, characterized in that: The method comprises: Initialize the algorithm parameters of MIWOA; wherein the algorithm parameters include the current number of iterations; When the current number of iterations does not reach the maximum number of iterations, obtaining a new shrinkage factor and an adaptive weight according to the current number of iterations and the maximum number of iterations; Obtaining an updated position update model according to the new shrinkage factor and the adaptive weight; wherein the position update model is obtained according to MIWOA; updating the position and corresponding fitness of each individual according to the updated position update model to obtain the updated position and corresponding fitness of each individual; wherein the individual includes a set of preset PID controller parameters; According to the updated positions and corresponding fitness of each individual, each individual is processed by a joiner mechanism of a sparrow search algorithm or a Cauchy-Gauss mutation strategy to obtain the best individual; A global optimal solution is obtained according to the position and fitness of the best individual; wherein the global optimal solution is the optimal PID controller parameters.

2. The PID parameter estimation method based on MIWOA according to claim 1, characterized in that: The new shrinkage factor is expressed as follows: Wherein, a represents the new shrinkage factor, ε is the first correlation coefficient, η is the second correlation coefficient, k represents the current iteration number, k max Represents the maximum number of iterations.

3. The PID parameter estimation method based on MIWOA according to claim 2, characterized in that: The adaptive weight is expressed as follows: Wherein, ω represents the adaptive weight, e is a natural constant, α is used to adjust the change of the adaptive weight vertically, and β is used to adjust the change of the adaptive weight horizontally.

4. The PID parameter estimation method based on MIWOA according to claim 1, characterized in that: The position update model includes three search mechanisms: surrounding prey, bubble net attack and searching prey.

5. The PID parameter estimation method based on MIWOA according to claim 4, characterized in that: The updating of the position and the corresponding fitness of each individual according to the updated position updating model to obtain the updated position and the corresponding fitness of each individual includes: Get the first random number; When the first random number is greater than or equal to a preset threshold, the updated positions and corresponding fitness of each individual are obtained according to the bubble network attack; Get the adjustment coefficient; When the first random number is less than the preset threshold value and the adjustment coefficient is greater than or equal to the adjustment coefficient, the updated position and corresponding fitness of each individual are obtained according to the search for prey; When the first random number is smaller than the preset threshold value, and the adjustment coefficient is smaller than the adjustment coefficient, the updated positions of each individual and the corresponding fitness are obtained according to the surrounded prey.

6. The PID parameter estimation method based on MIWOA according to claim 4, characterized in that: The encirclement of prey refers to the following formula: X(k+1)=ω·X * (k)-A·D1; Where X(k+1) represents the position and corresponding fitness of each individual after the k+1th iteration, k represents the number of iterations, ω is the adaptive weight, and X * (k) represents the current best individual position vector, D1 = |C·X * (k)-X(k)| represents the distance between other individuals and the current best individual, || represents the absolute value, A is the first coefficient vector, C is the second coefficient vector, and X(k) represents the position of each individual after the kth iteration and the corresponding fitness.

7. The PID parameter estimation method based on MIWOA according to claim 6, characterized in that: The bubble net attack refers to the following formula: X(k+1)=ω·X * (k)+D2·e bl ·cos(2πl); Where D2 = |X * (k)-X(k)| represents the distance between other individuals and the current best individual, b is a preset constant, l is a random number between [-1,1], and e is a natural constant.

8. The PID parameter estimation method based on MIWOA according to claim 7, characterized in that: The search for prey refers to the following formula: X(k+1)=ω·X rand (k)-A·D3; Among them, X rand (k) represents the position vector of a random individual randomly selected from the current individuals, D3 = |C·X rand (k)-X(k)| represents the distance between other individuals and the random individual.

9. The PID parameter estimation method based on MIWOA according to claim 1, characterized in that: According to the updated positions and corresponding fitness of each individual, each individual is processed by a sparrow search algorithm joiner mechanism or a Cauchy-Gauss mutation strategy to obtain the best individual, including: Get a second random number; When the second random number is less than the disturbance probability, each of the individuals is processed by the joiner mechanism of the sparrow search algorithm; When the second random number is greater than or equal to the disturbance probability, each individual is processed by the Cauchy-Gaussian mutation strategy.

10. The PID parameter estimation method based on MIWOA according to claim 9, characterized in that: When the second random number is less than the disturbance probability, processing each individual through the joiner mechanism of the sparrow search algorithm includes: When the second random number is less than the disturbance probability, generating a third random number; When the third random number is less than the safety threshold, each of the individuals is processed by the first formula of the joiner mechanism of the sparrow search algorithm; When the third random number is greater than or equal to the security threshold, each of the individuals is processed by the second formula of the joiner mechanism of the sparrow search algorithm.

Citation Information

Cited By

  • Self-adaptive PID (proportion integration differentiation) control method of gas turbine for load shedding working condition

    CN120990753A

  • Temperature control method and system based on IWOA optimization fuzzy PID

    CN121187393A