摘要
按照常规的设计方法,模糊控制系统的规则数量会随着输入变量数目的增加而呈指数性的增长。针对这一特点,引入了伸缩因子这一特征参数,研究了一种伸缩式三角形模糊集,根据系统不同控制区的灵敏度差异,通过改变模糊集的伸缩因子,动态调整模糊集的疏密度,从而灵活改变模糊集的划分;并在同样论域范围内,一定程度地减少规则数量,避免"规则爆炸"现象,提高系统的运行效率。将该模糊集划分方法运用于倒车入库的模糊控制中,其仿真结果表明了该方法的有效性。
According to conventional design method, rule quantity of fuzzy control system increases greatly with "the increasing of input variables. Aiming at the problem, telescopic factor is introduced, and a kind of telescopic triangle fuzzy set is studied. According to the difference of sensitivity in different control areas of the system, the telescopic factor of the fuzzy sets i.s changed and the density of the sets is regulated to change the division of the sets. In the same domain, the rule quantity is reduced to some extent to avoid "rules explosion", so as to increase the efficiency of the system. The proposed division method of fuzzy sets is applied to the fuzzy control of reversing a car into the garage, and the simulation result shows the effectiveness of the proposed method..
出处
《控制工程》
CSCD
2007年第B05期41-43,56,共4页
Control Engineering of China
关键词
规则爆炸
万能逼近定理
伸缩式三角形模糊集
伸缩因子
神经网络
rules explosion
almighty approaching theorem
telescopic triangle fuzzy sets
telescopic factor
neural network