航空学报 > 2016, Vol. 37 Issue (4): 1261-1271   doi: 10.7527/S1000-6893.2015.0160

一种快速鲁棒自适应故障估计方法设计

李飞, 赵国荣, 胡正高   

  1. 海军航空工程学院控制工程系, 烟台 264001
  • 收稿日期:2015-04-09 修回日期:2015-05-29 出版日期:2016-04-15 发布日期:2015-06-03
  • 通讯作者: 李飞,Tel.:0535-6635673 E-mail:xunzhaogancao@163.com E-mail:xunzhaogancao@163.com
  • 作者简介:李飞,男,博士,助教。主要研究方向:故障诊断与容错控制,机器视觉。Tel.:0535-6635673 E-mail:xunzhaogancao@163.com;赵国荣,男,博士,教授,博士生导师。主要研究方向:导航、制导与控制,故障诊断。Tel.:0535-6635673 E-mail:GRZhao6881@163.com;胡正高,男,博士研究生。主要研究方向:故障诊断与容错控制,微分方程。Tel.:0535-6635673 E-mail:zghu001@163.com
  • 基金资助:

    国家自然科学基金(61473306)

A design of fast robust adaptive fault estimation method

LI Fei, ZHAO Guorong, HU Zhenggao   

  1. Department of Control Engineering, Naval Aeronautical Engineering Institute College, Yantai 264001, China
  • Received:2015-04-09 Revised:2015-05-29 Online:2016-04-15 Published:2015-06-03
  • Supported by:

    National Natural Science Foundation of China (61473306)

摘要:

针对一类线性系统的执行器故障,对一种快速鲁棒自适应故障估计方法进行了设计。所设计的自适应故障估计观测器在放松严格正实(SPR)条件的同时,还具备对干扰的鲁棒性。在设计观测器参数时,根据有界实引理,对干扰和故障变化率的影响分别设计不同的约束不等式,能够更灵活地调节二者对故障估计影响的比例,并且引入松弛变量设计不同的Lyapunov矩阵,降低了设计的保守性。在设计过程中,由于出现了非线性矩阵不等式,本文采用了锥补线性化的算法进行求解,最终求得了最优的观测器参数。最后通过对某飞行器控制系统的仿真证明了算法的有效性。

关键词: 故障估计, 自适应, 观测器, 多约束, 锥补线性化

Abstract:

A fast robust adaptive fault estimation method is designed to reconstruct actuator fault of linear system. The strict positive realness (SPR) condition is relaxed by the adaptive fault estimation observer which simultaneously possesses robustness to disturbances. According to the bounded real lemma, different inequalities about disturbance and the derivative of fault are designed, so both the impact on fault estimation can be freely regulated. With slack variable, different Lyapunov matrices are designed to guarantee less conservatism. The cone complementarity linearization algorithm is utilized to solve the nonlinear matrix inequalities and the optimal observer parameters are obtained subsequently. The simulation results illustrate the effectiveness of the proposed method.

Key words: fault estimation, adaptive, observer, multi-constrained, cone complementarity linearization

中图分类号: