摘要
提出了一种针对纯滞后对象的神经网络及控制结构。在神经控制器的在线训练中，引入了反向传播的预测误差信号，并采用基于Ｋａｌｍａｎ滤波的学习算法分别训练用于辨识的神经网络及神经控制器。给出了具体的理论推证及运行机理分析。仿真结果表明，这种控制方案对纯滞后线性、非线性动态对象均能获得满意的控制效果。
A neural network based on structure for identification and control of plant with pure time delay is pressented.The predictive error signal and Kalmaa filter based learning algorithm are introduced in the process of on line train with neuarl controller.The theoretical proof and realization steps are discribed in detail.The potentials of the proposed structure are demonstrated by simulation method.The simulation results show that the controller is avaiable for controlling linear and nonlinear plant with pure time delay.
出处
《石油大学学报（自然科学版）》
CSCD
1998年第5期96-99,共4页
Journal of the University of Petroleum,China(Edition of Natural Science)
关键词
纯滞
神经网络
系统辨识
自适应控制
工业控制
pure time delay
neural network
systme identification
abaptive control
