A residual based adaptive unscented Kalman filter for fault recovery in attitude determination system of microsatellites

This paper presents an adaptive unscented Kalman filter (AUKF) to recover the satellite attitude in a fault detection and diagnosis (FDD) subsystem of microsatellites. The FDD subsystem includes a filter and an estimator with residual generators, hypothesis tests for fault detections and a reference logic table for fault isolations and fault recovery. The recovery process is based on the monitoring of mean and variance values of each attitude sensor behaviors from residual vectors. In the case of normal work, the residual vectors should be in the form of Gaussian white noise with zero mean and fixed variance. When the hypothesis tests for the residual vectors detect something unusual by comparing the mean and variance values with dynamic thresholds, the AUKF with real-time updated measurement noise covariance matrix will be used to recover the sensor faults. The scheme developed in this paper resolves the problem of the heavy and complex calculations during residual generations and therefore the delay in the isolation process is reduced. The numerical simulations for TSUBAME, a demonstration microsatellite of Tokyo Institute of Technology, are conducted and analyzed to demonstrate the working of the AUKF and FDD subsystem.     

叶片裂纹故障的多普勒在线监测技术仿真研究

本文研究了文献[1]介绍的叶片裂纹故障声学多普勒在线监测技术的监测原理;提出了一种新的基于裂纹叶片振动响应波形不对称性的监测方法;编制了这两种方法的计算及信号处理程序。仿真计算证明这些程序是可行的,同时表明,本文建立的新的监测方法在监测有效性、可靠性方面具有优势。