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.
出处
《自动化仪表》
CAS
北大核心
2010年第10期52-56,共5页
Process Automation Instrumentation
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
Fuzzy neural network control Robust control Monitor Electro-hydraulic servo system Real-time performance
