Abstract:

This paper proposes an energy management strategy optimized by dynamic programming for a series-parallel hybrid electric vehicle. The strategy involves modeling key powertrain system components and correcting model accuracy with experimental data. The dynamic programming algorithm is used to find the optimal control benchmark and analyze the mapping relationship between control variables and driving conditions. The study also investigates the influence of engine warm-up process on fuel consumption. The optimal operating mode distribution of clutches and engine and motor power are extracted from optimal control datasets. This study explores optimal operating characteristics of powertrain components for fuel saving and enables the strategy conversion from global optimal to real-time application, allowing for approximate optimal fuel economy in real-time for hybrid electric vehicles.