Capacitor Condition Monitoring Based on an Adaptive Observer of the Low-Frequency Capacitor Voltage Ripples for Modular Multilevel Converters

Abstract

Capacitor condition monitoring in modular multilevel converters (MMCs) has become a must in some critical industry applications. This is because capacitors represent a vulnerable component in power electronic converters, and MMCs usually integrate hundreds of them. This paper presents a new submodule (SM) capacitor condition monitoring method for MMCs based on an adaptive observer of the low-frequency capacitor voltage ripple. The proposed method does not require any additional hardware cost, it is intuitively simple, and computationally efficient. Moreover, it can provide capacitor condition monitoring of all the SMs in the converter simultaneously without affecting its operation. The operation principle of the proposed method consists in using the error between the measured SM capacitor voltage and its estimated value to derive the parametric uncertainty between the practical system SM capacitance and its nominal value assumed in the observer model. Therefore, if the SM capacitor exhibits a variation in its capacitance value, the proposed algorithm processes the consequent error signal to determine the actual capacitance. Subsequently, the observer model is updated, which brings the estimation error towards zero in the steady state. Simulation results on a 800-W MMC confirm the effectiveness of the proposed method.