基于反馈学习粒子群算法的极值搜索控制

Extremum Seeking Control Based on Feedback Learning Particle Swarm Optimization Algorithm

将基于反馈学习的粒子群(Feedback Learning Particle Swarm Optimization,FLPSO)算法引入极值搜索控制中,并且应用经典跟踪参考信号的方法,进一步改善极值搜索控制的性能.仿真结果显示,算法使系统控制输出平稳,并且系统性能输出快速渐进收敛到最优值,改善了基于格拉姆矩阵设计的极值搜索控制算法中存在的输出震荡问题.

This paper brought the feedback particle swarm optimization algorithm（FLPSO） into the extremum seeking control（ESC）,applied the idea of tracking problem which was first introduced into ESC by Zhang,and improved the performance of ESC.According to the simulation,the output of the control becomes relatively stable and the output of the performance function converges to optimum rapidly.The algorithm improves the problem of output function oscillation existed in the algorithm combined with Gramm method.