神经模糊逆模/PID复合控制在CSTR中的应用  

Composite Control Combining Neuro-Fuzzy Inverse Control with PID Control and Its Application in a Continuous Stirred Tank Reactor

作　　者：刘士荣[1] 俞金寿[2] 

机构地区：[1]宁波大学自动化系,宁波315211 [2]华东理工大学自动化研究所,上海200237

出　　处：《控制理论与应用》2001年第5期769-773,共5页Control Theory & Applications

基　　金：国家自然科学基金(69870 411)资助项目

摘　　要：研究了基于广义基函数神经模糊模型的逆系统实现及其直接逆模控制 ,并提出将直接逆模控制与PID反馈控制相结合的复合控制策略 .该控制策略已应用于CSTR的反应浓度控制 .仿真结果表明 ,神经模糊逆模 /PID复合控制能克服因辨识逆模型不精确引起的缺陷 ,并具有良好控制性能 .This paper studies the realization of inverse systems and direct inverse control strategies based on the generalized basis function neural fuzzy network (GBFNFN), and presents the composite control strategy combining neuro fuzzy inverse control with PID control. The composite control strategy has been applied to the control of the concentration of a continuous stirred tank reactor (CSTR). Simulation results show that the composite control strategy can overcome the defects of direct inverse control caused by inaccuracy of the identified inverse model, and that the system can have good control performances.

关 键 词：神经模糊模型 逆系统 逆模控制 PID控制 化学反应器 复合控制 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]