Design and Optimization of the Arm Inductor for Modular Multilevel Converter

The arm inductors utilized in a modular multilevel converter (MMC) can suppress the MMC circulating current, fault current, and output current, and circulating current ripples. This paper presents a four-step design and optimization method for the MMC arm inductor. The first step is the define the requirements and specifications of the arm inductor based on the MMC ratings and detailed modeling, in order to define the core material and select the appropriate winding wire as a second step. The third step is to develop a multi-objective optimization for the core design of the arm inductor, while the considered constraints are area product, maximum magnetic flux density, and window area, which is here done based on a Non-dominated Sorted Genetic Algorithm (NSGA-II). The objective functions are inductor total mass, total volume, total power losses, total core length, and air gap length. Finally, using the average ranking (AR) method, the best optimal solution is selected from the Pareto solution provided by NSGA-II.