PARAMETER IDENTIFICATION IN A HIGH PERFORMANCE HYDROSTATIC ACTUATION SYSTEM USING THE UNSCENTED KALMAN FILTER

This paper describes an early fault detection strategy for a high performance hydrostatic actuation system, referred to as the ElectroHydraulic actuator (EHA). Safety is crucial for the EHA which is being applied in flight surface actuation systems and in robotics. The proposed fault detection methodology in this manuscript uses a new state/parameter estimation algorithm, referred to as the Unscented Kalman Filter (UKF) to estimate parameters which cannot be measured using sensors. The parameters reflect the health condition of the system and changes in their normal values can be related to the inception and progression of faults in the system. The two parameters of interest in this study are the viscous damping coefficient of a symmetrical actuator and the effective bulk modulus of the hydrostatic system. The feasibility of the approach is demonstrated by a simulation study and using experimental data. Changes in the viscous damping coefficient provide valuable information about the lubricating properties of the oil and the seal conditions of the actuator. Changes in the effective bulk modulus, as a result of air getting trapped in the system, will change the system response, affecting the natural frequency and may cause stability problems. In this paper, the UKF is used for the first time for parameter estimation in a hydraulic system.