Home
Scholarly Works
An Analytical Method to Optimize the Power...
Journal article

An Analytical Method to Optimize the Power Allocation of Parallel Converters for Maximizing Efficiency Across All Load Levels

Abstract

For high-power conversion applications such as electric vehicle (EV) fast-chargers and wind power generation, multiple power electronic converters are often connected in parallel to reach high power levels with a modular design. To save energy and ease thermal stress, it is essential to maximize the total system efficiency over the wide range of power demands such applications experience. Most conventional power allocation methods share the power equally among the active modules, though some research has shown that unequal power sharing can lead to the highest total efficiency at some operating points. However, to this point, no general analytical method has been proposed to find the optimal power split of parallel converters (with equal or unequal power sharing) due to the complexity of the optimization problem. However, based on empirical observations from numerous simulations (up to 8 converters in parallel and 100 unique cases), this research has uncovered a key novel finding that reduces this multi-dimensional complex optimization problem to a 1-dimentional problem: at the optimal power split, out of countless possible power split possibilities between the parallel converters, only 1 or 2 non-zero power levels are operated at. This paper proposes a new analytical power allocation method for parallel converters based on this finding, which is based on the efficiency profile of a single converter. Simulation results for a variety of efficiency curves show efficiency increases up to 0.88%. Experimental results of 3 parallel boost converters show efficiency improvements up to 0.39%.

Authors

Sadi MH; Bauman J

Journal

IEEE Access, Vol. 14, , pp. 146619–146632

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Publication Date

January 1, 2026

DOI

10.1109/access.2026.3736307

ISSN

2169-3536

View published work (Non-McMaster Users)
An Analytical Method to Optimize the Power...