纯滞后系统的逆动态模型复合控制  

The Inverse Dynamic Model Compound Control for Dead Time System

作　　者：潘玉民[1] 永智群[1] 林亭生[1] 

机构地区：[1]华北科技学院,北京101601

出　　处：《自动化仪表》2005年第4期37-39,共3页Process Automation Instrumentation

摘　　要：提出了一种基于小脑模型神经网络的纯滞后对象控制方案。该方案通过在线学习被控对象的逆动态模型,利用前馈控制原理实现快速跟踪控制,克服了传统的Smith控制方法的不足。该方法具有不需建立精确模型、响应快速和鲁棒性强等特点。仿真结果表明控制器具有优良的品质。Based on CMAC neural network, a control strategy for dead time object is proposed. The strategy realized fast tracking control by using feed-forwarded controlling principle and on-line learning of object's inverse dynamic model, and overcome the shortage of traditional Smith control method. By this way, it is unnecessary to build up the accurate model. It offers quick response and high robustness. The results of simulation show that the controller is in excellent quality.

关 键 词：动态模型 纯滞后系统 复合控制 小脑模型神经网络 纯滞后对象 SMITH 控制方案 被控对象 在线学习 跟踪控制 控制原理 控制方法 精确模型 仿真结果 鲁棒性 前馈 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] F0[自动化与计算机技术—控制科学与工程]