电液伺服系统的模糊神经网络鲁棒控制  

Robust Control of Fuzzy Neural Network for Electro-hydraulic Servo System

作　　者：刘坤[1] 高少平[2] 张福军[1] 

机构地区：[1]黑龙江八一农垦大学信息技术学院,黑龙江大庆163319 [2]大庆石化公司机械厂,黑龙江大庆163711

出　　处：《自动化仪表》2010年第10期52-56,共5页Process Automation Instrumentation

摘　　要：为解决传统自适应控制模型的在线辨识和控制器的在线设计问题,基于模糊神经网络结构,提出了一种复合式控制方案。该方案通过引入运行监控器,克服了模糊神经网络实时性差的问题,实现了对不确定非线性系统的高精度输出跟踪;同时利用鲁棒反馈控制器,保证了模糊神经网络学习初期闭环系统的稳定性。该方案在电液伺服系统的应用中获得了满意的控制效果。In order to solve the problems of on-line identification of the traditional self-tuning control model and on-line design of the controller,the hybrid control scheme based on fuzzy neural network structure is proposed.In this control scheme,through introducing monitor,the problem of poor real-time performance of fuzzy neural network can be overcome;the high precision output tracking for indeterminist and non-linear system is realized.In addition,by adopting robust feedback controller,the stability of the closed-loop system in initial stage of fuzzy neural network learning is also guaranteed.The scheme has been applied into electro-hydraulic servo system with satisfactory control results.

关 键 词：模糊神经网络控制 鲁棒控制 监控器 电液伺服系统 实时性 

分 类 号：TP273.4[自动化与计算机技术—检测技术与自动化装置]