A Comprehensive Review of Intelligent Fault Diagnosis and Resilient Fault‐Tolerant Control in the Design of Modern Modular Multilevel Converters

ABSTRACT The modular multilevel converter (MMC) stands out as a potential power converter technology that is widely used for medium to high‐voltage converters, offering the benefits of modularity, scalability, and impressive harmonic performance. These benefits led to a widespread adoption in applications like high‐voltage direct current (HVDC), renewable energy systems, battery energy storage systems, and industrial drives. The MMC complex design is highly susceptible to faults, including open and short circuit faults in the SM, and arm voltage imbalances. The reliable and consistent operation of MMC under fault conditions is critical for operating in high‐voltage applications. This review paper presents an updated review of the MMC system fault mechanism, intelligent fault diagnosis technique, and resilient fault‐tolerant control strategies. This review paper systematically categorizes the various faults that occur in the MMC and provides an analysis of their impact on the converter's performance. The recent advancements in model‐based and data‐driven fault diagnosis techniques are critically analyzed, which include recent signal processing techniques and AI‐based methods, such as the integration of Physics‐Informed Neural Networks (PINNs) for MMC‐based applications. Furthermore, the advanced FTC‐based method and control strategies, including active, passive, and hybrid methods, are discussed in the context of MMC application. The paper also outlines the future research directions, which include digital twin‐based monitoring, scalable data‐driven control, and real‐time diagnosis for the next‐generation resilient MMC systems.