A-ECMS for Hybrid Heavy-Duty Vehicles

The equivalent consumption minimization strategy with real-time application prospects has attracted widespread attention. This paper proposes a two-layer hierarchical adaptive equivalent consumption minimum strategy for energy optimization of hybrid heavy-duty vehicles. In the upper layer, a feedforward neural network is employed for the purpose of forecasting both long-term and short-term velocity. The combination of historical velocity data and long-term velocity forecast data allows for the online identification of road conditions and the subsequent output of the corresponding battery state of charge set value. At last, the iterative update of the equivalence factor is concluded by taking into account the discrepancy between the battery state of charge set value and the current value, as well as the short-term velocity forecast data. The lower layer combines the equivalence factor output from the upper layer to complete the solution of the instantaneous equivalent fuel consumption minimisation problem, with engine torque acting as the control quantity. The simulation results show that the energy management strategy has good fuel economy and reduces fuel consumption by 5.2% compared with the traditional equivalent consumption minimum strategy.