Energy Management of Hybrid Electric Vehicles by DP-MPC-Based Mixed-Integer Nonlinear Control

This paper studies the optimization of energy management strategies (EMS) in hybrid electric vehicles (HEVs) to enhance fuel efficiency. While HEVs offer significant energy savings through coordinated use of power batteries and internal combustion engines, their full potential depends on effective EMS. The paper identifies key challenges in EMS, such as the coupling of discrete mode switching and continuous energy variables, leading to mixed-integer optimal control problems (MIOCP). Current optimization methods face computational and adaptability limitations, often resulting in suboptimal performance. To tackle these issues, the paper proposes a Dynamic Programming-Model Predictive Control (DP-MPC) framework. This approach combines Dynamic Programming’s global search capabilities for optimal mode switching with MPC’s receding horizon mechanism. Unlike conventional methods, this framework directly utilizes nonlinear physical models, avoiding simplifications that could lead to inaccuracies. Simulation results demonstrate the effectiveness of this approach in improving energy efficiency.