Hybrid Power Systems Simulation and Optimization Utilizing SSP and FMI

Abstract

Collaborative model-based development of the hybrid power system often requires large-scale co-simulation and system parameter optimization. In this study, we investigate an architecture for parallel processing simulation of SSP (System Structure and Parametrization) and FMI (Functional Mock-up Interface), which enables high-speed computation by multi-core distribution. We combine Bayesian optimization and co-simulation, then we build a collaborative development platform for hybrid power systems design. We report performance experiments using hybrid electric vehicle simulation model published by JAMBE (Japan Automotive Model-Based Engineering center).