This article presents the application of the additive form of the unscented Kalman filter (UKF-A) for the detection and isolation of incipient sensor faults in an electromechanical flight control actuation system. The detailed model of the actuator including mounting structure stiffness, load dynamics, compensators, power amplifier, and friction non-linearity is considered. The scheme utilizes the analytical redundancy that exists in the system between the linear actuator position, motor shaft angular velocity, and motor current for the diagnosis of incipient sensor faults in the system. Fault diagnosis is carried out using a single UKF-A, which is driven by the motor shaft velocity sensor. The onset of incipient sensor faults such as bias and scale factor faults in position and current sensor are detected and isolated using structured residuals that are generated using the UKF-A estimates. Bias and scale factor errors in the velocity sensor are detected using the statistical properties of the innovation sequence of the UKF-A. In the event of fault in the position sensor, real time reconfiguration using an estimate of the position from the UKF-A is used for continued operation of the system. Simulation results indicate that the isolation scheme formulated is capable of efficiently identifying incipient sensor faults and the reconfiguration scheme permits continued operation of the system in the presence of position sensor faults.
This paper presents a scheme for fault diagnosis in a flight control actuation system. The electromechanical control actuator considered here is based on a DC torque motor. The scheme utilizes the analytical redundancy that exists in the system between the linear actuator position, motor shaft angular velocity and motor current for diagnosis of incipient sensor faults in the system. Fault diagnosis is done using a single Kalman filter, which is driven by the motor shaft velocity sensor. Diagnosis of bias and scale factor faults in the position and current sensor is carried out using structured residuals that are generated using the Kalman filter estimates. Bias and scale factor errors in the velocity sensor are detected using the statistical properties of the innovation sequence of the Kalman filter. In the event of fault in the position sensor, real time reconfiguration using an estimate of the position from the Kalman filter is used for continued operation of the system. Robustness of the scheme to parameter variations is also examined.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.