PID control parameter optimization method
By improving the particle swarm optimization algorithm to optimize PID parameters, the problem of insufficient dynamic performance and robustness of existing PID controllers in complex systems is solved, achieving efficient and accurate parameter tuning and improving the performance of the control system.
Patent Information
- Application Number
- CN202511881794.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-15
- Publication Date
- 2026-02-10
AI Technical Summary
Existing PID control parameter tuning methods are difficult to balance the dynamic performance and robustness of complex systems, and cannot meet the needs of modern industry for high-precision, adaptive control.
An improved particle swarm optimization algorithm is adopted, which optimizes PID parameters by adjusting inertia weights and introducing fitness functions, thereby improving convergence speed and accuracy and achieving global optimal design.
This approach optimizes the performance of the PID controller, improves the efficiency and accuracy of parameter tuning, and meets the high-performance requirements of modern industrial control systems.
Smart Images
Figure CN121500731A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to PID control parameter tuning methods, and particularly to a PID parameter tuning method using an improved particle swarm optimization algorithm. Background Technology
[0002] Proportional-Integral-Derivative (PID) control, with its advantages of simple structure, clear physical meaning, strong robustness, and convenient engineering implementation, has become a core control strategy in the field of industrial automation. It is widely used in various linear and nonlinear control systems such as temperature, pressure, flow regulation, motor speed control, and robot control. The control performance of a PID controller directly depends on the parameter matching degree of the proportional coefficient, integral time constant, and derivative time constant. Therefore, parameter tuning is a key step in ensuring that the system achieves a balance between fast response, steady-state accuracy, and stability.
[0003] For a long time, industrial sites have mainly relied on traditional tuning techniques such as empirical trial and error and Ziegler-Nichols (ZN) method. These traditional methods are difficult to balance the dynamic performance and robustness of complex systems and can no longer meet the needs of modern industry for high-precision and adaptive control. There is an urgent need for a parameter tuning method to meet the high-performance requirements of modern industrial control systems. Summary of the Invention
[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose a PID parameter tuning method. This method draws on the particle swarm optimization algorithm and gives full play to its advantages of simple design, easy implementation, fast convergence speed and strong versatility. At the same time, the algorithm is improved to increase the convergence speed, and finally the globally optimal PID parameters can be obtained.
[0005] This invention adopts the following technical solution: a PID control parameter optimization method, comprising the following steps:
[0006] (1) Using PID control parameters as design variables, randomly generate them within the variable space enclosed by the upper and lower bounds of the design variables. A population composed of individuals;
[0007] (2) Obtain individual fitness based on the control performance indicators of the controlled equipment;
[0008] (3) Reorder the population according to individual fitness, and update the individual velocity and position according to the order;
[0009] (4) Update the individual with the worst fitness;
[0010] (5) When the maximum number of iterations is satisfied, the update is terminated and the optimal solution is output as the tuning parameter; otherwise, proceed to step (2).
[0011] The process involves using PID control parameters as design variables and randomly generating them within the variable space bounded by the upper and lower bounds of these design variables. The population consists of individuals, specifically as follows: the design variables are the proportionality coefficient P, the integral time constant I, and the differential time constant D; The individual's ID number; the individual's location. It is a vector consisting of design variables. The number of design variables.
[0012] In step (1), the population size The initial population is generated as follows:
[0013]
[0014] in, It is a random vector, and each element takes values in the range of 1. Random numbers between, This is the initial lower bound of the population. This is the initial upper bound of the population.
[0015] In step (2), the individual fitness The calculation method is as follows:
[0016]
[0017]
[0018]
[0019] in, , and These are the overshoot, settling time, and normalized steady-state error performance indicators, respectively. , and These are the weights of the system overshoot, settling time, and normalized performance index of steady-state error, respectively. , and Let be the steady-state error, overshoot time, and overshoot amount of the i-th individual in the t-th generation population, respectively. , and These represent the steady-state error, overshoot time, and maximum overshoot value. , and This is the adjustment coefficient.
[0020] In step (3), the individual update strategy incorporates information from the previous generation of individuals to enhance search capabilities.
[0021] In step (3), the update strategy is as follows;
[0022]
[0023]
[0024] in, For inertial weights, , , The acceleration factor represents the corresponding part. , , for Random numbers between, The i-th individual with the best fitness during the population iteration process is selected. The individual with the best fitness in the subgroup. The individual with the best fitness in the population; This represents the position of the i-th individual at iteration t. Let be the velocity of the i-th individual at iteration t.
[0025] In step (3), the inertia weight w decreases linearly with the number of iterations, and its calculation method is as follows:
[0026]
[0027] in, and And the set maximum and minimum weights, The maximum number of iterations, This represents the current iteration number.
[0028] In step (3), the acceleration factor is adjusted according to the number of iterations, and its calculation method is as follows:
[0029]
[0030] in, The maximum number of iterations, This represents the current iteration number.
[0031] In step (4), the worst-fit individual is updated, and the update steps are as follows:
[0032] The individual with the worst fitness in each subgroup is counted as... The update strategy is as follows:
[0033]
[0034] In the formula, , for Random numbers between, It is the individual with the best fitness in the population.
[0035] A PID control parameter optimization system, comprising:
[0036] The initialization module is used to randomly generate variables within a variable space bounded by the upper and lower bounds of the PID control parameters, using the PID control parameters as design variables. A population composed of individuals;
[0037] The fitness calculation module is used to obtain the individual fitness based on the control performance indicators of the controlled equipment;
[0038] The individual update module is used to reorder the population according to individual fitness, update the individual velocity and position based on the order, and update the worst-fit individuals.
[0039] The result output module is used to terminate the update when the maximum number of iterations is met, and output the optimal solution as the tuning parameters.
[0040] The advantages of this invention compared to the prior art are:
[0041] 1. This invention provides a new approach to PID parameter tuning. Based on the study of the shortcomings of the standard particle swarm optimization algorithm, the algorithm is improved by adjusting the inertia weight and introducing a fitness function, resulting in faster convergence speed, easier attainment of the global optimal design, and effective improvement of the accuracy of the optimal solution.
[0042] 2. This invention can optimize the performance indicators of PID controllers, improve the efficiency and accuracy of parameter tuning, and better meet the automatic control needs of electrical equipment. Attached Figure Description
[0043] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention. Detailed Implementation
[0044] The present invention will now be described in detail with reference to the accompanying drawings.
[0045] like Figure 1 As shown, this invention proposes a PID parameter tuning method that can be used for PID parameter tuning of various devices. Taking the PID control parameter tuning of an unmanned aerial vehicle (UAV) control system as an example, the specific implementation method is described, including the following steps:
[0046] (1) Using PID control parameters as design variables, establish constraints based on the controlled equipment; randomly generate within the variable space enclosed by the upper and lower bounds of the design variables. A population composed of individuals The individual's ID number; the individual's location. It is a vector consisting of design variables. This represents the number of design variables. , , This is the proportionality coefficient. The integral time constant is... is the differential time constant.
[0047] (2) Based on the control performance indicators of the controlled equipment, obtain the individual fitness. + , , and These are the system overshoot, settling time, and normalized performance indicators of steady-state error, respectively. , and These are the weights of the system overshoot, settling time, and normalized performance index of steady-state error, respectively; the specific calculation method for individual fitness is as follows;
[0048]
[0049]
[0050]
[0051] in, , and Let be the steady-state error, overshoot time, and overshoot amount of the i-th individual in the t-th generation population, respectively. , and This refers to the allowable steady-state error, overshoot time, and maximum overshoot value of the control system, which includes the controlled object. , and This is the adjustment factor. (Example) , , , , , .
[0052] (3) Reorder the population according to individual fitness, update individual velocity and position according to the order, and group them;
[0053] The update strategy is as follows;
[0054]
[0055]
[0056] The inertial weight decreases linearly with the number of iterations, and its calculation method is as follows:
[0057]
[0058] in, and And the set maximum and minimum weights, The maximum number of iterations, Current iteration number;
[0059] , , The acceleration factor, representing the corresponding part, is adjusted according to the number of iterations, and its calculation method is as follows:
[0060]
[0061] , , for Random numbers between, The i-th individual with the best fitness during the population iteration process is selected. The individual with the best fitness in the subgroup. It is the individual with the best fitness in the population.
[0062] (4) Update the individual with the worst fitness. The update steps are as follows:
[0063] The individual with the worst fitness in each subgroup is counted as... Its fitness score is The update strategy is as follows:
[0064]
[0065] In the formula , for Random numbers between intervals.
[0066] (5) Termination condition judgment. If the maximum number of iterations is met, the update is terminated and the optimal solution is output as the tuning parameter; otherwise, proceed to step (2).
[0067] In summary, this invention proposes a PID control parameter optimization method, constructs a simple and effective fitness function, and designs parameters such as inertia weight and acceleration factor that can change with the number of iterations by improving the particle swarm update strategy, so as to give it strong dynamic adjustment capability, coordinate and balance global search and local development, and improve the performance of the algorithm.
[0068] The above are merely specific steps of the present invention and do not constitute any limitation on the scope of protection of the present invention; it can be extended to the field of UAV operating condition optimization, and all technical solutions formed by equivalent transformation or equivalent substitution fall within the scope of protection of the present invention.
Claims
1. A method for optimizing PID control parameters, characterized in that, The steps include the following: (1) Using PID control parameters as design variables, randomly generate them within the variable space enclosed by the upper and lower bounds of the design variables. A population composed of individuals; (2) Obtain individual fitness based on the control performance indicators of the controlled equipment; (3) Reorder the population according to individual fitness, and update the individual velocity and position according to the order; (4) Update the individual with the worst fitness; (5) When the maximum number of iterations is satisfied, the update is terminated and the optimal solution is output as the tuning parameter; otherwise, proceed to step (2).
2. The PID control parameter optimization method according to claim 1, characterized in that: The process involves using PID control parameters as design variables and randomly generating them within the variable space bounded by the upper and lower bounds of these design variables. The population consists of individuals, specifically as follows: the design variables are the proportionality coefficient P, the integral time constant I, and the differential time constant D; The individual's ID number; the individual's location. It is a vector consisting of design variables. The number of design variables.
3. The PID control parameter optimization method according to claim 1, characterized in that, In step (1), the population size The initial population is generated as follows: ； in, It is a random vector, and each element takes values in the range of 1. Random numbers between, This is the initial lower bound of the population. This is the initial upper bound of the population.
4. The PID control parameter optimization method according to claim 1, characterized in that: In step (2), the individual fitness The calculation method is as follows: ； ； ； in, , and These are the overshoot, settling time, and normalized steady-state error performance indicators, respectively. , and These are the weights of the system overshoot, settling time, and normalized performance index of steady-state error, respectively. , and Let be the steady-state error, overshoot time, and overshoot amount of the i-th individual in the t-th generation population, respectively. , and These represent the steady-state error, overshoot time, and maximum overshoot value. , and This is the adjustment coefficient.
5. The PID control parameter optimization method according to claim 1, characterized in that: In step (3), the individual update strategy incorporates information from the previous generation of individuals to enhance search capabilities.
6. The PID control parameter optimization method according to claim 1, characterized in that: In step (3), the update strategy is as follows; ； ； in, For inertial weights, , , The acceleration factor represents the corresponding part. , , for Random numbers between, The i-th individual with the best fitness during the population iteration process is selected. The individual with the best fitness in the subgroup. The individual with the best fitness in the population; This represents the position of the i-th individual at iteration t. Let be the velocity of the i-th individual at iteration t.
7. The PID control parameter optimization method according to claim 6, characterized in that: In step (3), the inertia weight w decreases linearly with the number of iterations, and its calculation method is as follows: ； in, and And the set maximum and minimum weights, The maximum number of iterations, This represents the current iteration number.
8. The PID control parameter optimization method according to claim 6, characterized in that: In step (3), the acceleration factor is adjusted according to the number of iterations, and its calculation method is as follows: ； in, The maximum number of iterations, This represents the current iteration number.
9. The PID control parameter optimization method according to claim 1, characterized in that: In step (4), the worst-fit individual is updated, and the update steps are as follows: The individual with the worst fitness in each subgroup is counted as... The update strategy is as follows: ； In the formula, , for Random numbers between, It is the individual with the best fitness in the population.
10. A PID control parameter optimization system, characterized in that, include: The initialization module is used to randomly generate variables within a variable space bounded by the upper and lower bounds of the PID control parameters, using the PID control parameters as design variables. A population composed of individuals; The fitness calculation module is used to obtain the individual fitness based on the control performance indicators of the controlled equipment; The individual update module is used to reorder the population according to individual fitness, update the individual velocity and position based on the order, and update the worst-fit individuals. The result output module is used to terminate the update when the maximum number of iterations is met, and output the optimal solution as the tuning parameters.