Experimental design and genetic algorithm optimization of a fuzzy-logic supervisor for embedded electrical power systems

The embedded power systems are nowadays developing at high pace. Hybrid-electric vehicles, full-electric vehicles, airplanes, ships, high-speed trains, all share a common point – the embedded electrical power system. This paper aims to present an optimization methodology of a fuzzy-logic supervision strategy. The optimization objectives are to minimize the DC-link voltage variations, and to increase the system efficiency by reducing the dissipated power. For that, a methodology involving the experimental design and genetic algorithm will be presented. The simulation and experimental results are validating the proposed procedure.