摘要
本文提出了一种基于前馈神经网络IMC的直流电动机智能控制器的设计方法,其设计过程分为两步进行:第一步,训练一个神经网描述对象响应;第二步,训练一个网络描述对象的逆,并将此网络作为IMC控制器,并给出了仿真结果。
In this paper, a method based on IMC principle and neural networks is presented to design the intelligient controller of DC Motor. We proposed a two step procedure for using neural networks directly within the IMC structure. The first step involves training a networt to represent the plant reponse. The second step is to train a second network to present the inverse of plant. Having obtained the inverse model this network is then used as the controller of IMC. The simulahon results have demonstrated that this kind of controller can be used in the control systems with uncertain mathematics model or unknown parameters.
出处
《电气传动自动化》
1998年第1期38-40,共3页
Electric Drive Automation
关键词
直流电机
电机
智能控制器
设计
internal model control artificial neural networks model indentification inverse model identification robustness