PID response time calibration method and system based on grey wolf optimization algorithm

Abstract

The invention discloses a PID response time calibration method and system based on a grey wolf optimization algorithm. The method comprises the following steps: constructing a grey wolf optimization model for optimizing PID parameters; the grey wolf optimization model is initialized; performing iterative optimization on the grey wolf optimization model based on a memory bank and a random local search method to obtain optimized PID parameters; and inputting the optimized PID parameters into a PID control system of the MFC. According to the technical scheme provided by the invention, PID parameter optimization is carried out by utilizing the improved grey wolf optimization algorithm, and the memory bank and the random local search algorithm are introduced, so that the optimization capability of the grey wolf optimization algorithm is improved, and the PID parameter optimal solution with stable system and shortest response time is obtained; a computer can be used for calculation iteration, automatic calibration of PID control is achieved, the calculation speed does not depend on the experience of workers, the calibration speed is far higher than that of manual calibration, the labor cost is reduced, and the working efficiency is improved.

Description

Technical Field

[0001] The invention relates to the field of gas mass flow controllers, and in particular to a PID response time calibration method and system based on a Grey Wolf optimization algorithm. Background Art

[0002] A gas mass flow controller (MFC) is a device used to accurately measure and control the mass flow of gas. The working principle of the device is based on fluid mechanics and heat transfer principles. The mass flow rate of the gas is calculated by measuring the pressure difference and temperature change when the gas flows through the pipeline through a mass flow sensor, and the gas flow rate is adjusted by controlling the valve to keep it within the set range. MFC plays an important role in many fields, including semiconductor manufacturing, chemical engineering, biomedicine, aerospace, and environmental testing. These fields usually have certain requirements on the response time of the gas to meet production needs.

[0003] Proportional-integral-derivative control (PID) is widely used in industrial process control, including MFC response time calibration, due to its simple algorithm, good robustness and high reliability. When using the PID algorithm to calibrate the response time of MFC, it is necessary to manually adjust the PID parameters (including proportional coefficient, integral coefficient, and differential coefficient) according to the response curve phenomenon.

[0004] In actual PID parameter adjustment, the following problems often exist: (1) If the proportional coefficient is too large, the system's dynamic control capability will be reduced, or even unstable and cause downtime. (2) If the integral coefficient is too large, the control system will overshoot and delay the response time. (3) The differential coefficient can increase system stability, but it is too sensitive to external interference and will reduce the control system's anti-interference ability. Therefore, when using the PID algorithm for response time calibration, it is necessary to reasonably adjust the three PID parameters, and at the same time have the experience of observing the response curve to analyze and judge the parameters, and invest a certain amount of time and manpower costs to obtain PID response time control performance that meets high production needs. Summary of the invention

[0005] In view of the above challenges existing in the current technology, the present invention provides a PID response time calibration method based on the Grey Wolf Optimization Algorithm, which uses the improved Grey Wolf Optimization Algorithm to optimize PID parameters, introduces a memory library and a random local search algorithm, improves the optimization capability of the Grey Wolf Optimization Algorithm, and obtains the optimal solution of PID parameters when the system is stable and the response time is shortest.

[0006] To achieve the above object, the embodiments of the present invention adopt the following technical solutions: A PID response time calibration method based on the Grey Wolf optimization algorithm comprises the following steps: Construct a gray wolf optimization model to optimize PID parameters; Initialize the gray wolf optimization model; Based on the memory bank and random local search method, the gray wolf optimization model is iteratively optimized to obtain the optimized PID parameters; The optimized PID parameters are input into the PID control system of the MFC.

[0007] According to one aspect of the present invention, the PID parameters include: a proportional coefficient, an integral coefficient, and a differential coefficient.

[0008] According to one aspect of the present invention, the initializing the gray wolf optimization model includes: Set the value range of PID parameters; Set the number of search agents and maximum number of iterations for Gray Wolf Optimization; Set the fitness function for gray wolf optimization; According to the value range of the PID parameter, the PID parameter of the initial search agent is randomly generated; The fitness value of the initial search agent is calculated according to the fitness function to determine the optimal solution wolf.

[0009] According to one aspect of the present invention, the setting of the fitness function of the gray wolf optimization includes: The fitness function is constructed based on the integral of the absolute value of the time-weighted error of the MFC system.

[0010] According to one aspect of the present invention, the iterative optimization of the gray wolf optimization model based on the memory bank and the random local search method to obtain the optimized PID parameters includes: Construct a memory library to store the best individual wolf data in each iteration, including PID parameter values ​​and fitness values; During iteration, the PID parameter value is randomly searched locally, and the fitness value of the PID parameter value at each search point and iteration number is calculated according to the fitness function. The minimum fitness value is recorded and saved, and the memory bank and optimal solution are updated; Iterate to find the optimal PID parameters until the maximum number of iterations is reached, and save the optimized PID parameters and the best fitness value.

[0011] According to one aspect of the present invention, the iterative optimization of the gray wolf optimization model based on the memory bank and the random local search method to obtain the optimized PID parameters includes: When the iteration starts, a certain number of individual wolf data with lower fitness values ​​of the initial search agent are selected and saved into the memory bank.

[0012] According to one aspect of the present invention, the random local search for PID parameter values ​​comprises: A sample wolf is randomly selected from the wolf pack with a lower fitness value in the memory bank; The wolf closest to the sample wolf is found among other wolves through Euclidean distance, and a local wolf is randomly generated for iterative calculation.

[0013] According to one aspect of the present invention, the random local search for PID parameter values ​​further comprises: Randomly select 2 sample wolves from the wolf pack with lower fitness values ​​in the memory bank; Using genetic algorithm, a local wolf is generated based on two sample wolves for iterative calculation.

[0014] According to one aspect of the present invention, the PID response time calibration method based on the Grey Wolf optimization algorithm further includes: Test whether the response time and control curve of MFC meet the production requirements. If not, optimize the PID parameters again until they meet the production requirements.

[0015] A PID response time calibration system based on the Grey Wolf optimization algorithm, based on the PID response time calibration method based on the Grey Wolf optimization algorithm as described above, comprises: A model building module is used to build a gray wolf optimization model for optimizing PID parameters; Initialization module, used to initialize the gray wolf optimization model; Iterative module, used to iteratively optimize the gray wolf optimization model based on the memory bank and random local search method to obtain optimized PID parameters; The control module is used to input the optimized PID parameters into the PID control system of the MFC.

[0016] Advantages of the present invention: The present invention provides a PID response time calibration method based on the Grey Wolf optimization algorithm. The improved Grey Wolf optimization algorithm is used to optimize the PID parameters. A memory library and a random local search algorithm are introduced to improve the optimization ability of the Grey Wolf optimization algorithm, and the optimal solution of the PID parameters when the system is stable and the response time is the shortest is obtained. A computer can be used for calculation iteration to realize automatic calibration of the PID control. The calculation speed does not depend on the experience of the staff, and the calibration speed is much faster than manual calibration, thereby reducing the labor cost and improving the work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 This is a flow chart of a PID response time calibration method based on the Grey Wolf optimization algorithm according to Embodiment 1 of the present invention; Figure 2 This is a flow chart of a PID response time calibration method based on the Grey Wolf Optimization Algorithm described in Example 2 of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Embodiment 1

[0021] like Figure 1 As shown, a PID response time calibration method based on the gray wolf optimization algorithm includes the following steps: S1: Construct a grey wolf optimization model to optimize PID parameters.

[0022] When using the PID algorithm to calibrate the response time of the MFC, it is necessary to manually adjust the PID parameters according to the response curve phenomenon.

[0023] Specifically, the PID parameters include: proportional coefficient, integral coefficient, and differential coefficient.

[0024] Grey Wolf Optimizer (GWO) is a new swarm intelligence algorithm that simulates the social hierarchy and hunting behavior of grey wolves. It achieves the goal of finding the optimal solution through the process of tracking, surrounding prey and hunting. The principle of the Grey Wolf Optimizer is simple, with few adjustable parameters and strong global search capability. The key steps to find the optimal PID parameters through the Grey Wolf Optimizer are as follows: (1) Social system In the gray wolf optimization algorithm, the social hierarchy is mainly divided into α, β, δ and ω wolves. Among them, the best candidate value solution is set to α wolf, the second best candidate value is β wolf, the third best candidate value is δ wolf, and the remaining best candidate value is ω wolf. During the hunting process, the position of the wolf is updated according to the rules. At the same time, during the iteration process, the leading α, β, and δ gray wolves will be replaced by gray wolves with better adaptability.

[0025] (2) Surrounding the prey When the wolf pack is surrounding the prey, it is necessary to determine the distance between the individual and the prey. Based on the distance, the position of the gray wolf is updated. The update formula is:

[0026] Among them, X p (t) is the position vector of the prey, X(t) is the position vector of the gray wolf; D is the distance between the gray wolf and the prey, which can be calculated using Euclidean distance, Manhattan distance, etc.; t is the current iteration number; coefficients A and C can be expressed as:

[0027] Where b is the control parameter, b=2-2(t / t max ), t max is the maximum number of iterations that linearly decreases from 2 to 0 during the iteration process, and r1 and r2 are random vectors in [0,1].

[0028] (3) Hunting When the wolves have completed the encirclement of the prey, α wolf guides β wolf and δ wolf to reduce the encirclement of the prey to achieve the purpose of hunting. The mathematical expression of this process is as follows:

[0029] In the iterative loop, the final position expression of the gray wolf is:

[0030] Among them, X α , X β , X δ They represent the current positions of α wolf, β wolf and δ wolf respectively; X1, X2 and X3 represent the step lengths of α wolf, β wolf and δ wolf in their respective directions respectively.

[0031] Finally, the gray wolf optimization algorithm outputs the position of wolf α as the optimal solution for PID parameters and the corresponding fitness value.

[0032] S2: Initialize the gray wolf optimization model.

[0033] Step S2 includes: S21: Set the value range of PID parameters.

[0034] According to the response time curve characteristics of MFC, set the proportional coefficient K p , integral coefficient K i , differential coefficient K d The value range of .

[0035] S22: Set the number of search agents and maximum number of iterations for Gray Wolf Optimization.

[0036] Specifically, the number of search agents can be set to 30 and the maximum number of iterations can be set to 20.

[0037] S23: Set the fitness function for gray wolf optimization.

[0038] Step S23 includes: The fitness function is constructed based on the integral of the absolute value of the time-weighted error of the MFC system.

[0039] The integral of the time-weighted absolute error (ITAE) is a performance indicator commonly used in control system performance evaluation, especially in the parameter tuning process of PID controllers. The ITAE indicator comprehensively reflects the stability and dynamic response of the system by combining the absolute value of the error and the integral of time. It can be used as a fitness function. The specific calculation formula is:

[0040] Where e(t) represents the error between the system output and the expected output; t is time.

[0041] The smaller the ITAE index is, the better the stability of the system is and the faster the dynamic response is. The ITAE index is insensitive to the initial deviation, but very sensitive to the later deviation. When the ITAE index is used to adjust the controller parameters, the initial deviation of the transition process is large, but as time goes by, the deviation will quickly decrease, which shows that the use of the ITAE index can reduce the adjustment time. Therefore, in practical applications, ITAE is an important performance evaluation indicator.

[0042] S24: randomly generating PID parameters of an initial search agent according to the value range of the PID parameters.

[0043] According to the number of search agents set in step S22, the initial search agents are randomly generated. For example, in this method, the number of search agents is set to 30, and 30 groups of PID parameters are randomly generated as the initial search agents.

[0044] S25: Calculate the fitness value of the initial search agent according to the fitness function and determine the optimal solution wolf.

[0045] According to the fitness function set in step S23, the fitness values ​​of each group of PID parameters initially generated are calculated, and the one with the smallest fitness value is taken as the optimal solution wolf, that is, the three groups of PID parameters with the smallest, second smallest and third smallest fitness values ​​are set as α wolf, β wolf and δ wolf respectively, and the other wolves are set as ω wolf.

[0046] S3: Based on the memory bank and random local search method, the gray wolf optimization model is iteratively optimized to obtain the optimized PID parameters.

[0047] Step S3 includes: S31: Build a memory library to save the better individual wolf data in each iteration process, including PID parameter values ​​and fitness values.

[0048] The memory bank is a memory function set on the basis of the original gray wolf optimization algorithm, which saves the better individual wolf data in each iteration, including its position, fitness value, etc. After each iteration, the wolf with better indicators is updated to the wolf pack information in the memory bank to provide data for random local search.

[0049] Preferably, step S3 comprises: When the iteration starts, a certain number of individual wolf data with lower fitness values ​​of the initial search agent are selected and saved into the memory bank.

[0050] S32: When performing iteration, randomly search for PID parameter values ​​locally, calculate the fitness values ​​of the PID parameter values ​​at each search point and iteration number according to the fitness function, record and save the minimum fitness value, and update the memory bank and the optimal solution.

[0051] Specifically, the random local search PID parameter value includes: A sample wolf is randomly selected from the wolf pack with a lower fitness value in the memory bank; The wolf closest to the sample wolf is found among other wolves through Euclidean distance, and a local wolf is randomly generated for iterative calculation.

[0052] In practical applications, random local search can perform local search on the first 50% of individual wolf groups with lower fitness values ​​in the memory bank, select a sample wolf, and then find the wolf closest to the sample wolf among other wolves through Euclidean distance, and randomly generate a local wolf. The calculation formula for the position of the local wolf is:

[0053] Among them, X t,j is the generated local wolf position; X n,jis the position of the individual wolf closest to the sample wolf; X i,j is the selected sample wolf position; c1 is the acceleration coefficient.

[0054] S33: Iteratively search for optimal PID parameters until a maximum number of iterations is reached, and save the optimized PID parameters and the best fitness value.

[0055] Using the improved gray wolf optimization algorithm, the final optimized PID parameters, that is, the optimized proportional coefficient K, are obtained. p , integral coefficient K i , differential coefficient K d .

[0056] S4: Input the optimized PID parameters into the PID control system of the MFC.

[0057] In practical applications, this method can be linked with the PID control system of the MFC to achieve automatic calibration of the response time of the PID control.

[0058] The beneficial effects of this embodiment are: this method adopts the improved Grey Wolf optimization algorithm to optimize PID parameters, introduces a memory library and a random local search algorithm, improves the optimization ability of the Grey Wolf optimization algorithm, and obtains the optimal solution of PID parameters when the system is stable and the response time is shortest; a computer can be used for calculation iteration to realize automatic calibration of PID control, the calculation speed does not depend on the experience of the staff, the calibration speed is much faster than manual calibration, reduces labor costs, and improves work efficiency.

[0059] Embodiment 2

[0060] like Figure 2 As shown, a PID response time calibration method based on the gray wolf optimization algorithm includes the following steps: S1: Construct a grey wolf optimization model to optimize PID parameters.

[0061] When using the PID algorithm to calibrate the response time of the MFC, it is necessary to manually adjust the PID parameters according to the response curve phenomenon.

[0062] Specifically, the PID parameters include: proportional coefficient, integral coefficient, and differential coefficient.

[0063] Grey Wolf Optimizer (GWO) is a new swarm intelligence algorithm that simulates the social hierarchy and hunting behavior of grey wolves. It achieves the goal of finding the optimal solution through the process of tracking, surrounding prey and hunting. The principle of the Grey Wolf Optimizer is simple, with few adjustable parameters and strong global search capability. The key steps to find the optimal PID parameters through the Grey Wolf Optimizer are as follows: (1) Social system In the gray wolf optimization algorithm, the social hierarchy is mainly divided into α, β, δ and ω wolves. Among them, the best candidate value solution is set to α wolf, the second best candidate value is β wolf, the third best candidate value is δ wolf, and the remaining best candidate value is ω wolf. During the hunting process, the position of the wolf is updated according to the rules. At the same time, during the iteration process, the leading α, β, and δ gray wolves will be replaced by gray wolves with better adaptability.

[0064] (2) Surrounding the prey When the wolf pack is surrounding the prey, it is necessary to determine the distance between the individual and the prey. Based on the distance, the position of the gray wolf is updated. The update formula is:

[0065] Among them, X p (t) is the position vector of the prey, X(t) is the position vector of the gray wolf; D is the distance between the gray wolf and the prey, which can be calculated using Euclidean distance, Manhattan distance, etc.; t is the current iteration number; coefficients A and C can be expressed as:

[0066] Where b is the control parameter, b=2-2(t / t max ), t max is the maximum number of iterations that linearly decreases from 2 to 0 during the iteration process, and r1 and r2 are random vectors in [0,1].

[0067] (3) Hunting When the wolves have completed the encirclement of the prey, α wolf guides β wolf and δ wolf to reduce the encirclement of the prey to achieve the purpose of hunting. The mathematical expression of this process is as follows:

[0068] In the iterative loop, the final position expression of the gray wolf is:

[0069] Among them, X α , X β , X δ They represent the current positions of α wolf, β wolf and δ wolf respectively; X1, X2 and X3 represent the step lengths of α wolf, β wolf and δ wolf in their respective directions respectively.

[0070] Finally, the gray wolf optimization algorithm outputs the position of wolf α as the optimal solution for PID parameters and the corresponding fitness value.

[0071] S2: Initialize the gray wolf optimization model.

[0072] Step S2 includes: S21: Set the value range of PID parameters.

[0073] According to the response time curve characteristics of MFC, set the proportional coefficient K p , integral coefficient K i , differential coefficient K d The value range of .

[0074] S22: Set the number of search agents and maximum number of iterations for Gray Wolf Optimization.

[0075] Specifically, the number of search agents can be set to 30 and the maximum number of iterations can be set to 20.

[0076] S23: Set the fitness function for gray wolf optimization.

[0077] Step S23 includes: The fitness function is constructed based on the integral of the absolute value of the time-weighted error of the MFC system.

[0078] The integral of the time-weighted absolute error (ITAE) is a performance indicator commonly used in control system performance evaluation, especially in the parameter tuning process of PID controllers. The ITAE indicator comprehensively reflects the stability and dynamic response of the system by combining the absolute value of the error and the integral of time. It can be used as a fitness function. The specific calculation formula is: Where e(t) represents the error between the system output and the expected output; t is time.

[0079] The smaller the ITAE index is, the better the stability of the system is and the faster the dynamic response is. The ITAE index is insensitive to the initial deviation, but very sensitive to the later deviation. When the ITAE index is used to adjust the controller parameters, the initial deviation of the transition process is large, but as time goes by, the deviation will quickly decrease, which shows that the use of the ITAE index can reduce the adjustment time. Therefore, in practical applications, ITAE is an important performance evaluation indicator.

[0080] S24: randomly generating PID parameters of an initial search agent according to the value range of the PID parameters.

[0081] According to the number of search agents set in step S22, the initial search agents are randomly generated. For example, in this method, the number of search agents is set to 30, and 30 groups of PID parameters are randomly generated as the initial search agents.

[0082] S25: Calculate the fitness value of the initial search agent according to the fitness function and determine the optimal solution wolf.

[0083] According to the fitness function set in step S23, the fitness values ​​of each group of PID parameters initially generated are calculated, and the one with the smallest fitness value is taken as the optimal solution wolf, that is, the three groups of PID parameters with the smallest, second smallest and third smallest fitness values ​​are set as α wolf, β wolf and δ wolf respectively, and the other wolves are set as ω wolf.

[0084] S3: Based on the memory bank and random local search method, the gray wolf optimization model is iteratively optimized to obtain the optimized PID parameters.

[0085] Step S3 includes: S31: Build a memory library to save the better individual wolf data in each iteration process, including PID parameter values ​​and fitness values.

[0086] The memory bank is a memory function set on the basis of the original gray wolf optimization algorithm, which saves the better individual wolf data in each iteration, including its position, fitness value, etc. After each iteration, the wolf with better indicators is updated to the wolf pack information in the memory bank to provide data for random local search.

[0087] Preferably, step S3 comprises: When the iteration starts, a certain number of individual wolf data with lower fitness values ​​of the initial search agent are selected and saved into the memory bank.

[0088] S32: When performing iteration, randomly search for PID parameter values ​​locally, calculate the fitness values ​​of the PID parameter values ​​at each search point and iteration number according to the fitness function, record and save the minimum fitness value, and update the memory bank and the optimal solution.

[0089] Specifically, the random local search PID parameter value includes: Randomly select 2 sample wolves from the wolf pack with lower fitness values ​​in the memory bank; Using genetic algorithm, a local wolf is generated based on two sample wolves for iterative calculation.

[0090] In practical applications, crossover and mutation can be used to generate new local wolves. Crossover is to exchange part of the data of two sample wolves to generate new local wolves. Crossover operation is the main way to generate new solutions in genetic algorithms. Mutation is to make random changes to the new local wolves with a small probability to increase the diversity of the population. Mutation operation helps the algorithm jump out of the local optimal solution and explore a wider solution space.

[0091] S33: Iteratively search for optimal PID parameters until a maximum number of iterations is reached, and save the optimized PID parameters and the best fitness value.

[0092] Using the improved gray wolf optimization algorithm, the final optimized PID parameters, that is, the optimized proportional coefficient K, are obtained. p , integral coefficient K i , differential coefficient K d .

[0093] S4: Input the optimized PID parameters into the PID control system of the MFC.

[0094] In practical applications, this method can be linked with the PID control system of the MFC to achieve automatic calibration of the response time of the PID control.

[0095] Preferably, the method further comprises: S5: Test whether the response time and control curve of MFC meet the production requirements. If not, optimize the PID parameters again until they meet the production requirements.

[0096] The beneficial effects of this embodiment are: this method also uses a genetic algorithm to perform random local search, which helps the algorithm to jump out of the local optimal solution; and the PID parameters can be optimized multiple times according to actual production requirements until the requirements are met.

[0097] Embodiment 3

[0098] A PID response time calibration system based on the gray wolf optimization algorithm, based on the PID response time calibration method based on the gray wolf optimization algorithm as described in embodiment one or two, comprising: A model building module is used to build a gray wolf optimization model for optimizing PID parameters; Initialization module, used to initialize the gray wolf optimization model; Iterative module, used to iteratively optimize the gray wolf optimization model based on the memory bank and random local search method to obtain optimized PID parameters; The control module is used to input the optimized PID parameters into the PID control system of the MFC.

[0099] Preferably, the system further comprises: The optimization module is used to test whether the response time and control curve of the MFC meet the production requirements. If not, the PID parameters are optimized again until they meet the production requirements.

[0100] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the art within the technical scope disclosed in the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A PID response time calibration method based on the Grey Wolf optimization algorithm, characterized in that: The following steps are involved: Construct a gray wolf optimization model to optimize PID parameters; Initialize the gray wolf optimization model; Based on the memory bank and random local search method, the gray wolf optimization model is iteratively optimized to obtain the optimized PID parameters; The optimized PID parameters are input into the PID control system of the MFC.

2. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 1 is characterized in that: The PID parameters include: proportional coefficient, integral coefficient, and differential coefficient.

3. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 1 is characterized in that: Initializing the gray wolf optimization model includes: Set the value range of PID parameters; Set the number of search agents and maximum number of iterations for Gray Wolf Optimization; Set the fitness function for gray wolf optimization; According to the value range of the PID parameter, the PID parameter of the initial search agent is randomly generated; The fitness value of the initial search agent is calculated according to the fitness function to determine the optimal solution wolf.

4. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 3 is characterized in that: The fitness function of setting gray wolf optimization includes: The fitness function is constructed based on the integral of the absolute value of the time-weighted error of the MFC system.

5. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 3 is characterized in that: The gray wolf optimization model is iteratively optimized based on the memory bank and random local search method to obtain the optimized PID parameters including: Construct a memory library to store the best individual wolf data in each iteration, including PID parameter values ​​and fitness values; During iteration, the PID parameter value is randomly searched locally, and the fitness value of the PID parameter value at each search point and iteration number is calculated according to the fitness function. The minimum fitness value is recorded and saved, and the memory bank and optimal solution are updated; Iterate to find the optimal PID parameters until the maximum number of iterations is reached, and save the optimized PID parameters and the best fitness value.

6. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 5 is characterized in that: The gray wolf optimization model is iteratively optimized based on the memory bank and random local search method to obtain the optimized PID parameters including: When the iteration starts, a certain number of individual wolf data with lower fitness values ​​of the initial search agent are selected and saved into the memory bank.

7. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 5 is characterized in that: The random local search PID parameter value includes: A sample wolf is randomly selected from the wolf pack with a lower fitness value in the memory bank; The wolf closest to the sample wolf is found among other wolves through Euclidean distance, and a local wolf is randomly generated for iterative calculation.

8. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 5 is characterized in that: The random local search PID parameter value also includes: Randomly select 2 sample wolves from the wolf pack with lower fitness values ​​in the memory bank; Using genetic algorithm, a local wolf is generated based on two sample wolves for iterative calculation.

9. The PID response time calibration method based on the Grey Wolf optimization algorithm according to claim 1 is characterized in that: The PID response time calibration method based on the Grey Wolf optimization algorithm also includes: Test whether the response time and control curve of MFC meet the production requirements. If not, optimize the PID parameters again until they meet the production requirements.

10. A PID response time calibration system based on the Grey Wolf optimization algorithm, characterized in that: The PID response time calibration method based on the Grey Wolf optimization algorithm according to any one of claims 1 to 9 comprises: A model building module is used to build a gray wolf optimization model for optimizing PID parameters; Initialization module, used to initialize the gray wolf optimization model; Iterative module, used to iteratively optimize the gray wolf optimization model based on the memory bank and random local search method to obtain optimized PID parameters; The control module is used to input the optimized PID parameters into the PID control system of the MFC.

Images