Real‐Time Energy Management Strategy for Fuel Cell Plug‐In Hybrid Electric Bus Using Short‐Term Power Smoothing Prediction and Distance Adaptive State‐of‐Charge Consumption

Abstract

To improve energy consumption economy and fuel cell life economy, this article proposes a multiobjective real‐time energy management strategy based on short‐term power smoothing prediction and distance adaptive state‐of‐charge (SOC) consumption for fuel cell/battery plug‐in hybrid electric bus. First, a distance adaptive SOC real‐time reference trajectory planning method is proposed by collecting the actual operating condition data of Dalian city buses. Second, the motor power is smoothed using the first‐order exponential smoothing approach, and the long short term memory network is adopted to predict the smoothed motor power. Finally, based on the predicted motor power and a real‐time reference SOC, a model predictive control strategy is designed in the finite prediction horizon, and dynamic programming is used to solve multiobjective cost functions to achieve optimal power allocation. Simulation results show that the proposed strategy has remarkable performance in terms of the total operating cost, SOC sustenance, and fuel cell degradation compared with the rule‐based strategy. In addition, the average time of online energy distribution in each step of the proposed strategy is 43.8 ms, which is much smaller than the simulation step length of 1s.