Combined optimal sizing and energy management of hybrid electric vehicles

This paper describes a new methodology for sizing energy sources in hybrid electric vehicles, that enables obtaining the minimal sizing required for a given driving cycle, independently of the chosen energy management strategy. The methodology is based on two combined optimization loops: one for sizing the energy sources, using a genetic algorithm, and another one for computing the optimal energy management strategy for a specific driving cycle, using dynamic programming. Results show that the algorithm can find the best sizing of sources for the best fuel consumption, with a 6.5kW fuel cell and a 75Wh battery for the ECE driving cycle and a 9.0kW fuel cell and a 72Wh battery for the LA92 cycle. Compared to results obtained through the mean sizing power method, the algorithm shows that the hydrogen consumption can be reduced by up to 70% and the size of the battery by up to 67 %. The proposed methodology can thus help optimize the sizing of hybrid vehicles used for given driving cycles.