Applying the Extended Kalman Filter to Fault Diagnosis in the Control System of a Manned Carrier Rocket

Abstract

The fault diagnosis system of the control system is one of important subsystems for a manned carrier rocket which can alarm and start its escape program in case of any faults. Its key part is the fault detection algorithm. In the full paper, we explain in some detail the fault diagnosis system and the application of the extended Kalman filter (EKF) to the fault detection of the control system of the carrier rocket; in this abstract, we just add some pertinent remarks to naming the first two sections of the full paper. In section 1, we present the discrete EKF. In section 2, we establish the 3-DOF mathematical model of the carrier rocket and discretize the mathematical model and obtain the EKF model. We point out that the fault in the control system usually causes changes in the parameters of the control system. But the EKF is not robust to the changes and its residual errors increase or even disperse when the parameters suddenly change. The analysis of the amplitude of residual error can determine whether the control system has some faults. Finally we give a numerical simulation example to verify the effectiveness of the fault detection algorithm. The simulation results, shown in Figs. 1 and 2 in the full paper, indicate preliminarily that the state estimates of the EKF deviate from the measurement values and that the filtering residual error increases rapidly, indicating that there are some faults in the control system.