Research of the Method on the RBF Neural NetworksInversion Dynamic Control

摘要：

In accordance with the fact that uncertainty and disturbance factors occurred to the radar seeker control system during the missile flight and the demand of dynamic performance and tracking accuracy for the system, the method based on the RBF neural networks inversion dynamic control is presented. On the basis of the PID feedback control,the RBF neural networks inversion dynamic control structure is constructed by combining the RBF neural networks with the inversion dynamic control. The RBF neural networks inversion dynamic control strategies are presented and the corresponding model of the controller is established by adopting the learning algorithm of the orthogonal least squares based on the RBF neural networks. The digital simulation of the designed system is performed by using MATLB/Simulink. The simulation results show that the RBF neural networks inversion dynamic control is superior to the traditional PID control and can effectively improve the dynamic performance, steady-state performance and anti-disturbance performance of the system.

关键词： PID control RBF neural networks inversion dynamic control least squares learning algorithm simulation

作者: WU Xiaoyan FAN Zongchen WU Jing

作者单位: 25＃, Sanyuan 713800, Shaanxi Chian

会议类型: 国际会议

会议名称: 第八届国际测试技术研讨会(8th International Symposium on Test and Measurement)

会议地点: 重庆

会议语种:英文

页码: 1567-1570

在线出版日期: 2009-08-01（万方平台首次上网日期，不代表论文的发表时间）