模糊逻辑系统的一种自适应参数学习算法  

An Adaptive Parameters Learning Algorithm for Fuzzy Logic Systems

作　　者：祖家奎[1] 赵淳生[1] 戴冠中[2] 

机构地区：[1]南京航空航天大学超声电机研究中心,江苏南京210016 [2]西北工业大学自动控制系,陕西西安710072

出　　处：《系统仿真学报》2004年第5期1108-1110,共3页Journal of System Simulation

摘　　要：根据自适应模糊逻辑理论以及基于导数的参数优化方法,分析了模糊逻辑系统常规参数优化算法存在的不足,从改善模糊系统的参数训练步长的角度着手,对其常规的最速下降算法进行了改进,提出了一种新的自适应模糊逻辑系统参数学习算法,并对该算法的推导过程进行了具体的分析和描述。最后,针对一非线性函数逼近问题,对该算法进行了验证,仿真的逼近精度和收敛速度都获得了提高,表明本文提出的模糊逻辑系统自适应参数学习算法是可行和有效性的,且能够克服常规参数优化算法中存在的某些不足。According to the theories of adaptive fuzzy logic system and general parameter optimization, the characteristics of normal parameter-learning algorithms on fuzzy logic systems are firstly analyzed, and its some defects are discussed briefly. Aimed at improving parameter-training pace, a new adaptive parameter-learning algorithm is then advanced, which is deduced and explained in detail. Lastly, the proposed algorithm has been evaluated by a nonlinear function to approximate its ideal values; system抯 simulation precision and its rate of convergence are enhanced remarkably. Those results demonstrate the efficiency and feasibility of the proposed adaptive optimization algorithm, which can overcome some deficiencies of normal parameter-learning algorithms.

关 键 词：模糊逻辑系统 自适应 训练步长 参数学习算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]