Hopfield神经网络在极点配置自校正控制中的应用  

Application of Hopfield Neural Network in Pole Placement Self - tuning Control

作　　者：王科俊 金鸿章 李国斌 

出　　处：《哈尔滨工程大学学报》1995年第3期32-37,共6页Journal of Harbin Engineering University

基　　金：高等学校博士学科点专项科研基金

摘　　要：Ｈｏｐｆｉｅｌｄ神经网络在极点配置自校正控制中的应用王科俊，金鸿章，李国斌（自动控制系）［摘要］将Ｈｏｐｆｉｅｌｄ神经网络应用于极点配置间接自校正控制器设计，提出用一阶Ｔａｙｌｏｒ逼近神经元输出作为参数估计的方法。通过仿真试验研究了该方法的性能。得到...In this paper, an indirect self-tuning controller based on pole placement is designed withthe application of Hopfield neural network. The method of parameter estimation is proposedwith the first level Talor application. The characteristics of method proposed are evaluated bymeans of simulation. The results obtained are compared with those of exponentially weightedrecursive least square method in parameter estimation and Gaussian elimination method in solving Doophantine equation in order to highlight the effectiveness of the proposed control strategyby the Hopfield neural network.

关 键 词：神经网络 极点配置 参数估计 自校正控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]