摘要
针对模糊控制器的优化问题,提出了一种基于改进的分层遗传算法在线优化设计方法。该算法在分层遗传算法的基础上修改了信息交换方式,采用自适应交叉算子和变异算子,并改进了变异算子的变异方式,使其能在现有最优解基础上进行更精确的局部搜索,提高了搜索速度和精度;同时,使用了具有约束的时间与绝对误差乘积积分(ITAE)型性能指标函数,能够对系统的稳态误差、超调量和上升时间进行有侧重的优化;并结合最小二乘参数在线辨识技术,实现了时滞时变系统模糊控制器的参数和结构在线同步快速优化。仿真实验证明了该方法的有效性。
Aiming at the problem of optimization of fuzzy controllers, a method of online optimization design based on the improved hierarchic genetic algorithm is proposed. The method modifies the information exchange style on the basis of hierarchic genetic algorithm. Using and improving the adaptive crossover operator and mu- tation operator, the genetic algorithm has better local search ability based upon current optimal solutions, so the search speed and search precision of genetic algorithm are increased. Meanwhile, by using restricted integration performance index ITAE, the optimization can be emphasized particularly on the steady-state error, the overshoot or the rise time. By introducing the LS parameter online identification, the parameters and structure of fuzzy controllers can be fast optimized simultaneously in the time-delay and time-varying system. Simulation experiments illustrate the validity of the proposed method.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2009年第4期911-915,共5页
Systems Engineering and Electronics
基金
甘肃省自然科学基金(3ZS042-B25-039)
光电技术与智能控制教育部重点实验室(兰州交通大学)开放基金(k04106)
兰州市科技发展计划(2008-1-2)资助课题
关键词
分层遗传算法
模糊控制
隶属度函数
模糊规则
在线优化
hierarchic genetic algorithm
fuzzy control
membership function
fuzzy rule
online optimization
