一类工业过程的神经元智能控制  

Adaptive Neuron Control of a Kind of Industrial Processes

作　　者：王永骥[1] 徐桂英[1] 涂健[1] 

机构地区：[1]华中理工大学自动控制工程系

出　　处：《华中理工大学学报》1994年第12期51-55,共5页Journal of Huazhong University of Science and Technology

摘　　要：研究了基于自适应神经元的一类工业过程控制问题，分析了闭环系统稳定性与神经元控制器参数的关系，给出了闭环系统稳定的充分条件及神经元学习参数初值的一种方法及学习的收敛性证明。理论分析表明，神经元智能过程控制系统有很强的鲁棒性和适应性，仿真结果表明了其优良的性能。The controller design of akind of industrial processes based on theadaptive neuron isstudied. The relationship between neuron controller parameters and closed-loop control sta-hlitv of such a svstem is analvzed and a sufficient condition of stability and a method of de-termining the initial weighting coefficients are given,A theoretical analysis shows that theneuron process control system has strong robustness and adaptivity. Simulation resuIts showthat the performance of the neuron control system is satisfactory.

关 键 词：神经元 智能控制 神经网络 工业过程控制 

分 类 号：TP273.1[自动化与计算机技术—检测技术与自动化装置]