Rule-based improved MPC–SQP real-time energy management for hybrid electric heavy-duty trucks based on vehicle prediction

The energy management strategy (EMS) is crucial to the energy-saving potential of hybrid electric heavy-duty trucks. This study proposes a rule-based, improved model predictive control–sequential quadratic programming (MPC–SQP) EMS combined with a velocity prediction model. First, a velocity prediction model based on long short-term memory is introduced and its prediction capability is verified by a quantitative analysis. Second, the operating mode of hybrid electric heavy-duty truck is classified into six types by a rule-based EMS, and MPC with SQP solver is used for driving charging mode and hybrid driving mode. Then, comparative analysis among proposed rule-based improved MPC–SQP, dynamic programming (DP), model predictive control–dynamic programming (MPC–DP), and rule-based method is conducted on MATLAB/Simulink simulation platform. With the proposed EMS, vehicle fuel consumption is reduced by 16.2% and 14% compared to the rule-based method and MPC–DP, respectively, and achieves over 6.6% of the theoretical minimum fuel consumption obtained by DP. Finally, the engineering applicability of proposed EMS is validated through hardware-in-the-loop test.