基于鲁棒自适应观测器的线性系统故障估计  

Robust Adaptive Observer-Based Fault Estimation For Linear Systems

作　　者：王洲辉[1] 张科[1] 韩治国[1] 

机构地区：[1]西北工业大学航天学院,陕西西安710072

出　　处：《计算机仿真》2016年第7期131-134,140,共5页Computer Simulation

基　　金：国家自然科学基金(61174204;61101191;61502391)

摘　　要：针对刚体航天器在轨运行时存在执行机构失效故障和外部扰动问题,提出了一种采用鲁棒自适应观测器的故障估计方法。采用观测器不仅能够获得系统状态和执行器故障的渐进估计,而且具有广泛的适用范围。通过采用H"性能指标,降低系统扰动对故障估计精度的影响。在设计观测器增益矩阵时,为了改善故障估计的性能,引入了区域极点配置,通过把误差方程的极点配置在理想的稳定区域来改善故障估计性能,提高故障估计的准确性。最后,将改进方法应用到微小卫星姿态控制系统中,仿真验证了所提方法的有效性,为系统故障优化估计提供了科学依据。In this paper, a robust adaptive observer-based fault estimation method was developed to detect actuator faults of a spacecraft in the presence of external disturbances. This observer can acquire the asymptotic estimates of the original system state and the actuator fault term simultaneously, and has a broad scope of application. By using the H∞ performance, the effect of the system disturbance on the estimation accuracy of faults can be reduced. In the design of the observer gain matrix, in order to improve the performance of fanh estimation, regional pole assignment was introduced. Placing the poles of error equation in an ideal area can improve the performance and accuracy of fault estimation. Finally, this method was applied to the microsatellite attitude control systems and the simulations verify the effectiveness of the method.

关 键 词：鲁棒自适应 观测器设计 时变故障 故障估计 

分 类 号：V411.8[航空宇航科学与技术—航空宇航推进理论与工程]