摘要
在对一些复杂系统进行模糊控制时,由于对系统的不了解,很难得到合适的控制规则。基于模糊控制器的一种解析结构,提出了将模糊控制器与径向基函数(Radial Basis Functio)神经网络相结合的方法。由RBF神经网络对系统进行辨识,并为学习系统提供必要的信息,根据信息对经验规则进行修改,从而改善模糊控制系统动态响应。仿真结果表明该控制器对模型参数变化具有较好的适应能力,能够较快地修改系统的原控制规则,使对象输出较快地跟踪系统的输入。
When we apply fuzzy control to an unknown system,we can't find good control rules because we don't know its characteristics. Based on an analytic structure of fuzzy controller, a new method is put forward, which combines fuzzy controller and RBF (Radial Basis Function). Using RBF to make system identification and provide necessary message for learning system, and based on the message provided by RBF to modify experiences rules, the dynamic response of fuzzy control system can be improved. Simulation results show that this controller has better adaptability for the variety of model parameters and can modify ariginal system control rules, and makes the output follow the system input tracks faster.
出处
《电气自动化》
北大核心
2001年第1期16-18,共3页
Electrical Automation
