北京航空航天大学学报 ›› 2013, Vol. 39 ›› Issue (12): 1601-1606.

• 论文 • 上一篇    下一篇

基于RBF网络的航空发动机预测滑模控制

苗卓广, 谢寿生, 丁键, 王磊   

  1. 空军工程大学 航空航天工程学院, 西安 710038
  • 收稿日期:2013-01-22 出版日期:2013-12-30 发布日期:2013-12-27

Predictive sliding mode control for aero-engine based on RBF network

Miao Zhuoguang, Xie Shousheng, Ding Jian, Wang Lei   

  1. The Aeronautics and Astronautics Engineering College, Air Force Engineering University, Xi'an 710038, China
  • Received:2013-01-22 Online:2013-12-30 Published:2013-12-27

摘要: 针对航空发动机是一个不确定性的强非线性系统,借鉴预测控制的思想,提出了基于径向基函数RBF (Radical Basis Function)网络的航空发动机预测滑模控制.首先利用RBF网络建立航空发动机预测模型,进而得到滑模预测模型;其次在线修正网络参数实时反馈校正滑模预测模型,滚动优化求取控制量;然后采用另外一个RBF神经网络实现了全包线建模和控制;最后分析了控制系统的收敛性.仿真结果表明,所设计的控制器性能良好,能有效地抑制参数摄动和干扰的影响.

Abstract: A method of predictive sliding mode control based on RBF network was put forward for aero-engine with uncertainty and strong nonlinearity. The method drew lessons from the ideology of predictive control. Predictive modal of aero-engine was established by using RBF network and sliding mode predictive modal was deduced. Sliding mode predictive modal was modified online by adjusting network parameters. And the control value was get by making rolling optimization. Full flight envelop control was made using another RBF network. The stability condition was analyzed. Simulation results show that the devised controller has good effect and restrains the influence of parameters perturbation and interfere.

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发