PID Parameter Optimization Based On Mixture Genetic Algorithm And It's Application In Liquid-level Control

PID algorithm is widely used in industrial process control because it has simple structure and strong robustness. As a crucial factor in improving control performance, Parameters tuning of the PID controller has been studying. Genetic algorithm (GA) solves the optimization problem by simulating the natural processes and mechanisms of biological evolution. GA possesses some defects such as poor local search and slow convergence. Based on these above reasons, a hybrid genetic algorithm (HGA) has been designed to optimize the PID parameters.Firstly, based on the analysis for present situation and development of GA, a HGA is presented using optimal saving strategies, adaptive crossover and mutation probability, which make the result moving to the area of the best solution quickly. Moreover, combining simplex operator with GA, the HGA can improve the capability of local search and find the best solution quickly. And then, the HGA is employed to optimize the PID parameters. HGA can obtain optimal solution after much less iteration, reduce the number of evaluations of the fitness function and improve the search efficiency of the algorithm. The detailed simulation flowchart is given and the numerical simulations for different models of charged object under different control performance index are implemented. The simulation results illustrated the effectiveness of the presented algorithm.Finally, the presented algorithm is applied to single tank liquid-level system of the process control apparatus A3000. The system model is identified using transition curve method. The HGA and PID controller is compiled in Matlab language and the system identification and PID parameters optimization are naturally integrated. The effectiveness of the optimized PID parameters is demonstrated by the experiments and the experimental results are satisfactory.