Observer Based Fault Estimation for Inverter Devices of Traction Systems with Disturbance

This paper is concerned with the design of a fault estimation scheme with component fault and sensor noise for the three-phase inverter of high-speed trains. To design the component fault estimation observer, the three-phase inverter system states are augmented with the disturbance treated as new states, leading to an augmented and singular system. The observer is designed to generate the needed residual signal for fault estimation by using system output measurements. The adaptive law is established by the Lyapunov theory, and the design of the proposed observer is reformulated as linear matrix inequality (LMI). Simulation result shows the effectiveness of the proposed approach for fault estimation of the three-phase inverter.