The improved adaptive filter algorithm applied to sensor of control system for failure detection and identification

Using standard multiple model adaptive filter to detect faults, a bank of Kalman filters are needed to finish filtering computation. When a large number of kalman filters are used, MMAE can be a very costly algorithm. An improved filter algorithm for MMAE is developed that uses a single Kalman filter to produce equivalent residuals of a complete Kalman filter bank. The computation cost of the method is less than the stand Kalman filter bank algorithm.