Comparing Extended Kalman Filter and Particle Filter for estimating field and damper bar currents in Brushless Wound Field Synchronous Generator for stator winding fault detection and diagnosis

Condition monitoring of the Brushless Wound Field Synchronous Generator (BWFSG) is important as it is widely used in mission and safety critical applications such as marine vessels. The frequency signatures in rotor field current are well known indicators to detect and diagnose stator winding short circuits in synchronous generators, in addition the damper bars current could also be used for fault detection and diagnosis. However, BWFSG these currents are not accessible. Hence, it is important to use the mathematical model and state estimation techniques to estimate these parameters. This paper compares the performance of state estimation techniques such as the Extended Kalman Filter (EKF) and Particle Filter (PF) in estimating field current and damper bars current for stator winding fault detection and diagnosis. The experimental validation results confirmed that the performance of the EKF is better than that of the PF in terms of number of stator winding fault signatures extracted from estimating the field current and damper bars currents. Thus, the future work proposed to use the EKF for developing Model-Based Condition Monitoring (MBCM) system for the BWFSG.