Modeling and design of the modular multilevel converter with parametric and model-form uncertainty quantification

This paper presents a design methodology with uncertainty quantification to estimate the required margin for two main design variables in a modular multilevel converter (MMC). In this methodology, the minimum required design margins are calculated by quantifying all sources of uncertainty in the modeling and simulation of MMCs. To this end, an enhanced modeling framework is presented to take into account the parametric uncertainty (PU) that results from manufacturing variability, and model-form uncertainty (MFU) that results from inherent inaccuracies of the models used in the design process. In this paper, sensitivity analysis (SA) is used to guide the modeling effort and minimize the number of uncertain parameters required for inclusion in uncertainty quantification. Besides, a simplified testbed for model validation of the MMC is developed. This testbed is used for conducting low-power validation experiments and Monte-Carlo simulations to estimate PU and MFU, respectively.