摘要
针对高精度伺服系统中存在的各种非线性及不确定性 ,提出了基于模糊神经网络的复合控制方法。控制器由前馈控制器和闭环反馈控制器组成。前馈控制器由零相差跟踪控制器 (ZPETC)及FIR滤波器构成 ,闭环反馈控制器采用模糊神经网络控制 ,包含一个辨识网络FNNI和控制网络FNNC。该控制方法结合了神经网络和模糊推理的优点 ,可以克服各种非线性和不确定性 ,明显提高跟踪性能 ,具有很好的鲁棒性。提出的模糊神经网络结构简单 ,推理算法易于实现 ,并且可以更合理地选择初始权值 ,加快了网络的收敛速度 。
A digital tracking controller based on fuzzy-neural network for high precision servo systems is presented. This controller consists of feedforward controller and feedback controller. The feedforward controller is composed of a zero phase error tracking controller and a zero phase low-pass filter. The feedback controller is composed of two fuzzy-neural networks, one used for identification (FNNI) and the other for as controlling (FNNC). The proposed adaptive controller combines the advantages of neural network and fuzzy reasoning. Experimental results demonstrate that the proposed controller can effectively improve the tracking performance and is robust to parameter uncertainty and external disturbance. This FNNC presented in this paper has a simple structure and its reasoning algorithm is easily implemented.With this FNNC the initial parameter can be chosen more reasonably and the time of neural network learning can be reduced, thus, the real-time performance of neural network is improved.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2003年第8期980-984,共5页
Systems Engineering and Electronics
关键词
前馈
跟踪
模糊神经网络
鲁棒性
Feedforward
Tracking
Fuzzy-neural network
Robustness
