A data-driven real time energy management strategy for fuel cell hybrid vehicle in the connected environment.

Energy management of fuel cell hybrid vehicles (FCHVs) in the connected environment has attracted widespread attention. In this paper, a data-driven on-line equivalent consumption minimum strategy (ECMS) is proposed to study the influence of different term vehicle-to-everything (V2X) information on fuel economy. In the presented method, the relationship between the ECMS and Pontryagin’s minimum principle (PMP) is established firstly. Then a long short-term memory (LSTM) network is built which the input variables are related with vehicle states and features of different term V2X information. Finally, the trained LSTM network is applied online to tune the equivalent factor (EF) to achieve fuel economy. The simulation results has proven the proposed energy management strategy (EMS) can achieve near-optimal fuel economy and the V2X information can further enhance energy-saving potential of FCHVs.