Power Management Optimization of a Fuel Cell/Battery/Supercapacitor Hybrid System for Transit Bus Applications

In this paper, the optimization of a power management strategy of a fuel cell/battery/supercapacitor hybrid vehicular system is investigated, both offline and in real time. Two offline optimization algorithms, namely, dynamic programming and Pontryagin's minimum principle, are first compared. The offline optimum is used as a benchmark when designing a real-time strategy, which is an inevitable step since the offline optimum is not real-time capable and is oriented only toward minimizing hydrogen consumption, which may result in the unnecessary overloading of the battery. The design and optimization of the real-time strategy makes use of a multiobjective genetic algorithm while taking into account, apart from hydrogen consumption, other important factors, such as the slow dynamics of the fuel cell system and minimizing the battery power burden. As a result, the real-time strategy is found to consume slightly more hydrogen than the offline optimum; however, it dramatically improves system durability.