改进差分进化算法在PID控制器在线优化中的应用

Differential Evolution Used in the Online Optimization of PID Controller

将基于混合差分策略的改进差分进化优化算法应用在PID控制器在线优化中。MDE结合了DE/rand/1的多样性和DE/best/1的高收敛速度的优点,算法的寻优性能远超过两种策略单独作用时的性能。MDE充分利用了差分进化的进化寻优优势与PID控制器的简单方便,让控制器在系统运行过程中进行自我设计和优化。

This paper presents a new method——a modified differential evolution algorithms(MDE) for the the online optimization of PID Controller,where DE / rand / 1 and DE / best / 1 run currently.MDE unified the high convergence rate merit of the DE / rand / 1 and the DE / besl / l multiplicity, the algorithm optimization performance surpasses two kind of strategies affect when alone the performance far. MDE has used the difference evolution evolution optimization superiority and the PID controller fully simple convenient, lets the controller carry on the individual plan and program and the optimization in the systems operation process.