Online Model-Based Condition Monitoring for Brushless Wound-Field Synchronous Generator to Detect and Diagnose Stator Windings Turn-to-Turn Shorts Using Extended Kalman Filter

Abstract

In this paper, a model-based approach is proposed to detect and diagnose stator winding fault in the Brushless wound-field synchronous generator (BWFSG). The extended Kalman filter is used as a state and parameter estimation technique for the proposed model-based approach. The mathematical model of the BWFSG with stator winding fault is developed and simplified for online implementation. An experimental test-rig is used to acquire the required inputs for the developed state estimation technique. The estimated rotor currents and fault parameter are analyzed to identify key signatures for condition monitoring (CM). The harmonic components such as the second harmonic components of the estimated field and damper currents, and the rms value of the estimated fault parameters are identified as suitable signatures for winding fault and diagnose. Based on the identified signatures, a model-based CM algorithm is proposed and validated in real time. The validation results confirmed that the proposed algorithm is able to detect and diagnose winding inter-turn short-circuit faults in real-time reliably.