东北大学学报(自然科学版) ›› 2005, Vol. 26 ›› Issue (10): 919-922.DOI: -

• 论著 •    下一篇

基于神经网络的STATCOM非线性鲁棒逆推设计

刘恩东;井元伟;王珂;张嗣瀛   

  1. 东北大学信息科学与工程学院;东北大学信息科学与工程学院;辽宁电力有限公司;东北大学信息科学与工程学院 辽宁沈阳110004
  • 收稿日期:2013-06-24 修回日期:2013-06-24 出版日期:2005-10-15 发布日期:2013-06-24
  • 通讯作者: Liu, E.-D.
  • 作者简介:-
  • 基金资助:
    国家自然科学基金资助项目(60274009)

Nonlinear robust backstepping design of STATCOM using neural networks

Liu, En-Dong (1); Jing, Yuan-Wei (1); Wang, Ke (2); Zhang, Si-Ying (1)   

  1. (1) School of Information Science and Engineering, Northeastern University, Shenyang 110004, China; (2) Liaoning Electric Power Co. Ltd., Shenyang 110006, China
  • Received:2013-06-24 Revised:2013-06-24 Online:2005-10-15 Published:2013-06-24
  • Contact: Liu, E.-D.
  • About author:-
  • Supported by:
    -

摘要: 针对带有STATCOM的单机无穷大总线系统,提出了一种新的神经网络逆推控制方法,设计了STATCOM的非线性控制器.这个方法对于参数变动是系统化的和鲁棒的,且是根据众所周知的逆推控制技术.所得的控制器能够保证跟踪误差和连接权有界.同时不需要预先知道神经网络连接权最优值的上界.神经网络的连接权值在线进行调整,而无需离线学习.仿真结果表明该方法的有效性.

关键词: 电力系统, 柔性交流输电系统, 静止同步补偿器, 神经网络, 逆推

Abstract: A novel backstepping control technique based neural networks is presented for single machine infinite bus system with STATCOM to design a nonlinear controller. Motivated by the well-known backstepping design technique, the method is systematic and robust to parameter variation. The controller derived can guarantee the boundedness to both tracking error and weight updates. No prior knowledge of the upper bound on the optimal weight values is required. The weights of neural networks are updating on-line, and the off-line training phase is not required. Numerical simulation was carried out to illustrate and clarify the approach.

中图分类号: