Application of Multi-strategy Flower Pollination Algorithm under Exhaustive Search in PID Parameter Optimization
PID controller is widely used in the industrial field,with simple structure,excellent control effect,and parameters play an absolute role in the controller.Flower pollination algorithm is a widely used meta-heuristic algorithm,but it has some shortcomings,such as easy to fall into local optimum,slow convergence speed in the late iteration stage,and poor optimization accuracy.The original fixed probability is replaced by the conversion probability of the full search strategy of the antihyperbolic tangent function with random pertur-bations,and the global search and local search are balanced.The introduction of new dynamic factors in cross-pollination,the influence of maternal pollen position is changed.In the self-pollination stage,a weighted learning strategy is proposed to make the pollen gather in the direction of excellent pollen.Somersault exploration strategies are introduced to increase population diversity.On this basis,the multi-strategy flower pollination algorithm under exhaustive search(MSFPA)is proposed,and the simulation experiments of four algorithms,namely Flower Pollination Algorithm(FPA),Particle Swarm Optimization(PSO),Differential Evolution Algorithm(DE)and Elephant Swarm Algorithm(EHO),are compared and analyzed.It is showed that the MSFPA has fast convergence speed and better performance.The MSFPA algorithm is applied to the PID parameter optimization,and the optimized system has a small overshoot,a short adjustment time,and strong stability.
国家科技期刊平台
NSTL
万方数据
