Active fault diagnosis for hybrid systems based on sensitivity analysis and EKF

An active fault diagnosis (AFD) approach for different kinds of faults is proposed. The AFD approach excites the system by injecting a so-called excitation input. The input is designed off-line based on a sensitivity analysis in order that the maximum sensitivity for each individual system parameter is obtained. Using the maximum sensitivity results in better precision in the estimation of the corresponding parameter. The fault detection and isolation is done by comparing the nominal parameters with those estimated by an extended Kalman filter. In this study, Gaussian noise is used as the input disturbance as well as the measurement noise for simulation. This method is implemented on a large scale livestock hybrid ventilation model which was obtained during previous research.