摘要
采用将T-S模型与RBF神经网络相结合的网络结构,提出一种复合式控制方案,以解决传统自适应控制中模型的在线辨识和控制器的在线设计问题,以达到对不确定非线性系统的高精确度输出跟踪控制;通过引入运行监控器,克服模糊神经网络控制方法通常存在的实时性差的问题;同时,利用一个鲁棒反馈控制器,来保证模糊神经网络模型学习初期闭环系统的稳定性.并应用于电液位置伺服系统的仿真研究中,获得满意的控制效果.
In this paper,a kind of hybrid control structure is designed using the fuzzy neural of combining the T-S fuzzy model and the RBF neural theory to solve these problems which are about the on-line realization of model and the on-line design of controller in traditional adaptive control.The work above is to track the output of the uncertain unlinear system with high accuracy.Through using the supervisor,the common problem of poor real-time in fuzzy neural control is come over.With a robust feedback controller,the stability of closed-loop system in fuzzy neural model learning earlier is guaranteed.On the other hand,this paper utilizes this scheme to an electro-hydraulic position servo system simulation and achieves satisfactory results.
出处
《哈尔滨理工大学学报》
CAS
北大核心
2011年第1期69-74,共6页
Journal of Harbin University of Science and Technology
