Spectral analysis‐based fault diagnosis algorithm for 3‐phase passive rectifiers in renewable energy systems

Abstract

Three-phase passive (or uncontrolled) rectifiers are typically placed at the front end of the overall power conditioning circuit for AC to DC conversion in low-cost, small-scale renewable-energy systems, e.g. residential purposes wind turbines. Whenever a diode in one leg or multiple legs becomes faulty, offering an open circuit to the flow of current in both directions, it severely affects the power quality at the grid-side and may lead to the permanent disconnection of the power contributing unit from the grid. Even under a normal voltage rectification mode, an unbalance fault on the generator side may cause similar behaviour. To address these issues, this study presents a novel spectral analysis based fault detection and diagnosis (FDD) method for unbalance and open-circuited faults in the 3-phase uncontrolled rectifier. The proposed FDD algorithm accurately identifies the fault location based on magnitude and phase angles of different harmonic components of the filtered rectified voltage and specifically devised thresholds. The effectiveness of the proposed FDD algorithm is demonstrated through exhaustive MATLAB simulations of the grid-connected power conditioning unit with variable 3-phase supply. Subsequently, the hardware implementation of the overall experiment is demonstrated under a fixed voltage, and a fixed frequency source fed 3-phase uncontrolled rectifier.