提出一种基于数值微分的卫星陀螺故障诊断方法。通过引入代数可观测的概念,将故障诊断问题转化为求解数值微分问题。并根据卫星姿态运动学方程证明了陀螺故障的代数可观性。然后通过高增益观测器的方法来近似姿态敏感器测量输出的数值导数,并利用李雅普诺夫理论分析和设计高增益观测器。基于陀螺故障的代数可观性和星敏感器测量输出的数值微分,可以直接得到陀螺组件的故障估计值。最后,通过突变、缓变和并发故障等仿真算例验证了算法的有效性,仿真结果表明所提出的方法不仅可以检测故障的发生,而且能够估计故障幅值。
A novel fault diagnosis approach for gyroscopes in satellites is proposed in this paper. The fault diagnosis problem is transformed to calculation of numerical differentiation by using the concept of algebraic observable. Based on the satellite attitude kinematics equation, it is proved that the fault of gyroscopes is algebraic observable. And then the numerical derivatives of the measurement outputs of attitude sensors are approximated via high gain observers analyzed and designed by tile Lyapunov stability theory. Since the fault of gyroscopes is algebraic observable and the numerical derivative of the measurement of attitude sensors is available, the gyroscope faults can be estimated directly. Finally, numerical simulations for several fault scenarios are conducted to verify the effectiveness of the fault diagnosis scheme. Simulation results demonstrate that the presented method can not only detect the occurenee of faults, but also estimate their maanitudes.