论文部分内容阅读
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(改进差分进化算法)结合了差分策略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。此方法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。
The improved differential evolution algorithm based on hybrid difference strategy is applied to online optimization of PID controller. MDE (Improved Differential Evolution Algorithm) combines the advantages of the diversity of DE / rand / 1 and the high convergence rate of DE / best / 1. The performance of the algorithm outperforms the performance of the two strategies alone. This method takes full advantage of the evolutionary evolution advantage of differential evolution and the simple and convenient PID controller, allowing the controller to self-design and optimize during the operation of the system.