在线变结构模糊神经网络的自适应内模控制  

An Adaptive Internal Model Control Based on an Online Variable-Structure Fuzzy Neural Network

作　　者：曾道蓉[1] 赵曜[1] 

机构地区：[1]四川大学电气信息学院,四川成都610065

出　　处：《自动化技术与应用》2006年第11期7-9,48,共4页Techniques of Automation and Applications

摘　　要：本文提出一种在线变结构补偿模糊神经网络训练算法。该在线变结构算法,使得出的网络结构简单。并且由于该网络引入了补偿模糊神经元,使网络能对模糊规则进行在线的训练。将此算法应用到仿射非线性动态系统和大时滞线性动态系统的内模控制中。仿真结果表明,该方法能有效的控制动态过程,具有较好的自适应性和鲁棒性。An online variable-structure training algorithm for fuzzy compensation network is proposed. And a compensation neuron is used in the network for on-line training of the fuzzy rules. Then this algorithm is applied to an affine-nonlinear dynamic plant and in a linear dynamic plant with large time-delay. Simulation shows that the method is effective and has excellent properties of adaptability and robustness.

关 键 词：模糊神经网络 内模控制 非线性动态系统 最近邻聚类学习算法 

分 类 号：TP273.21[自动化与计算机技术—检测技术与自动化装置]