Detecting and Localizing Open-Circuit Switch Faults in MMCs Using a Model Informed Estimation Scheme With Low Computational Complexity

Abstract

The detection and localization of switch open-circuit faults (OCFs) in modular multilevel converters (MMCs) is crucial for enhancing their reliability. This article presents a model informed estimation-based fault detection and localization (FDL) scheme with low computational burden and implementation complexity. Its main novelty comprises two parts: derivation of a new model that quantifies the expected deviation in submodule capacitor voltages due to OCFs, and utilization of a Disturbance Observer (DOB) that, by leveraging the derived model, needs only one signature waveform for each arm. As a result, the proposed FDL scheme enables estimation of OCFs while maintaining very low and constant computational burden and implementation complexity regardless of the number of installed submodules per arm. To the best of the authors' knowledge, this work is the first to explore the use of OCFs models that can quantify the OCFs-induced deviations in the MMC capacitor voltages. Experimental results verify that the proposed model-informed FDL scheme with DOB can detect and localize OCFs accurately and rapidly, while retaining important traits such as robustness to load changes and measurement noise.