PROJECT TITLE :

Non-parametric fault detection methods in non-linear systems

ABSTRACT:

In this study, 2 fault detection methods are proposed for non-linear systems. In these methods, Gaussian process (GP) is integrated into extended Kalman filter (EKF) and square root cubature Kalman filter (SCKF), which are referred to as GP-EKF and GP-SCKF, respectively. The foremost necessary advantage of the proposed methods is that there's no would like to know the accurate model of the system. Thus, these methods are thought of as non-parametric approaches of fault detection in non-linear systems. Moreover, by applying these methods, it is attainable to detect the fault with high accuracy at early stage. 1st, GP-EKF and GP-SCKF are proposed for non-linear state estimation, and then GP-SCKF is compared with the GP-EKF and the results of this comparison prove the prevalence of GP-SCKF relating to the complexity of computations and accuracy. Further, simulation results show a good performance of GP-EKF and GP-SCKF in non-linear system's fault detection. To illustrate performance of these algorithms in state estimation and fault detection, they're used in aircraft tracking system. Both of the proposed ways can detect the sensors faults at early stage.