时滞系统控制参数优化的遗传模拟退火算法  

Genetic Simulated Annealing Algorithm of Controller Parameter Optimization for Time-delay Systems

作　　者：杨磊[1] 田作华[1] 刘山[1] 杨晓军[1] 

机构地区：[1]上海交通大学自动控制系,上海200030

出　　处：《计算机仿真》2006年第2期154-156,共3页Computer Simulation

摘　　要：该文基于遗传模拟退火算法,提出一种时滞系统的控制参数优化方法,同时对M atlab遗传算法工具箱GAOT进行改进,使之适用于PID参数的优化。该文所采用的算法保留了遗传算法和模拟退火算法分别在全局和局部搜索能力强的优点,能克服常规遗传算法中解的早熟现象、局部寻优能力差,难以保证对参数优化的计算效率和可靠性要求等缺陷。研究表明,改进后的遗传模拟退火算法是一种行之有效的方法,具有实用价值。In this paper, a controller parameter optimization method based on the genetic simulated annealing algorithm for time - delay system is proposed. GA toolbox - based MATLAB is simultaneously modified to apply to PID parameter optimization. Genetic simulated annealing algorithm is a new kind of global searching algorithm, which is based on the genetic algorithm and the simulated annealing algorithm. It can overcome a lot of GA＇s defects, such as premature convergence, bad local searching ability and GA deceptive problems. An optimization of PID parameter of a time - delay system is presented. Simulation results show this new strategy is effective.

关 键 词：宣传算法 模拟退火算法 时滞系统 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]