基于改进型遗传算法的九点控制器参数自整定方法  

Parameter self-tuning in nine-point controller ased on improved genetic algorithms

作　　者：刘教瑜[1] 杨艳[1] 陶亮[1] 

机构地区：[1]武汉理工大学自动化学院,湖北武汉430070

出　　处：《电力自动化设备》2005年第4期28-31,共4页Electric Power Automation Equipment

摘　　要：九点控制器是一种新型的基本逻辑控制器。重点介绍了九点控制器原理,即当被控对象偏离设定值时,九点控制器自动加入一个反作用力,迫使系统响应回归到设定值附近。改进型遗传算法在原有实数编码策略中加入了择优和判断终止操作;阐述了其选择、交叉、变异、择优、判断终止5种操作过程。给出了九点控制器系统的结构方框图及程序框图;随机产生20个种群,每个种群8个个体经过10代、20代、50代和100代的参数整定,系统性能对比表明:九点控制器的参数可以自整定,整定效果满足系统的约束条件。The nine-point controller is a new type of basic logic controller. Its principle is empha-sized,that is,when the controlled object deviates from the set value,a counterforce will be exerted to force it back. An improved genetic algorithm is introduced to the controller parameter tuning,in which optimal selection and stop judgment are added to real-coded strategy. The operations of selection,across,mutation,optimal selection and stop judgment are expounded. The block diagram and program flowchart of the nine-point controller system are offered. Simulations are carried out,in which 20 groups with 8 individuals in each group are optimized through 10,20,50 and 100 generations. The system performance contrast shows that the parameters of nine-point controller are self-tuned to satisfy the system constraint conditions.

关 键 词：九点控制器 遗传算法 参数自整定 

分 类 号：TM744[电气工程—电力系统及自动化] TM761