Adaptive Kalman filtering for sensor fault estimation and isolation of satellite attitude control based on descriptor systems

Sensor fault estimation and isolation is significant for an attitude control systems model of a satellite, as it works in a complex environment. The standard unscented Kalman filter algorithm may lose its accuracy when the noise is considerable. Therefore, an adaptive filtering algorithm is proposed based on the sampled-data descriptor model. The performance of the unscented Kalman filter in sensor fault estimation is improved by the adaptive algorithm depending on innovation and the measurement residual, and its convergence is guaranteed. Combining the adaptive unscented Kalman filter with the multiple-model adaptive estimation, a sensor fault isolation method is proposed. Finally, simulation examples show that this algorithm has better estimating accuracy and isolation results.