Abstract
KEYWORDS－parallel hybrid electric vehicle, stochastic dynamic programming, particle swarm optimization, hardware-in-the-loop test, energy management strategy
ABSTRACT－Hybrid electric vehicles (HEVs) are able to switch to the appropriate drive modes in different working conditions and achieve the purpose of improving fuel economy and reducing emissions simultaneously. Therefore, it offers an effective approach to dealing with current problems like exhaustion of oil resources and severe pollution of environment, etc. For the hybrid electric vehicle, its energy management strategy basically determines the power distribution between the engine and the battery pack, and naturally the strategy is one of the most important factors influencing the fuel economy. In this paper, the parallel HEV is selected for study with reference to the idea of stochastic dynamic programming (SDP). Driver’s demanded power in New European Driving Cycle (NEDC) was abstracted into a stochastic process that is changing with the vehicle speed. Meanwhile, particle swarm optimization (PSO) is adopted to optimize gear selection. Under the restriction of keeping the battery capacity constant and the gear efficiency sub-optimal, the overall energy management strategy for the vehicle is established to minimizing fuel consumption. When the strategy is applied into the Matlab/Simulink-based simulation of forward-facing vehicle model, it is demonstrated that compared with rule-based energy management strategy, the proposed strategy can reduce fuel consumption by 9.5% in the NEDC. In addition, the feasibility and validity of this strategy are verified through the hardware-in-the-loop (HIL) test.