摘要
针对 Takagi- Sugeno模糊逻辑系统的隶属函数不具有自适应性且模糊规则数的确定带有很大的人为主观性 ,这里引入了一类广义 Takagi- Sugeno模糊逻辑系统 ;在模型实现上 ,以广义 Takagi- Sugeno模型为个体 ,采用简单、有效的矩阵编码方式 ,借助遗传算法得到一个次优的广义 Takagi- Sugeno模糊系统模型 ,该模型不仅能很好地逼近所要辨识的非线性系统 ,而且还具有较低的复杂度 .仿真结果表明了广义 Takagi- Sugeno模型及其参数辨识方法的正确性和有效性 .
In Takagi-Sugeno fuzzy logical system, its membership functions have no self-adaptability and the number of fuzzy rules is defined subjectively. A generalized Takagi-Sugeno fuzzy logical system model is quoted. In search of optimal parameters of the generalized Takagi-Sugeno model the matrix coding is adopted. The structure of the generalized Takagi-Sugeno model is evolved by GA and the resulting suboptimal solution can be found quickly, which has lower complexity and approximates to a nonlinear system very well. The validity of this method is demonstrated by a numerical simulation.
出处
《自动化学报》
EI
CSCD
北大核心
2002年第4期581-586,共6页
Acta Automatica Sinica
基金
北京市自然科学基金 ( 4992 0 0 7)资助
关键词
遗传算法
广义Takagi-Sugeno
模糊逻辑系统
最优参数辨识
矩阵编码
模糊控制
Computer simulation
Encoding (symbols)
Fuzzy sets
Genetic algorithms
Identification (control systems)
Mathematical models
Matrix algebra
Membership functions
