Hierarchical model predictive control for energy management and lifespan protection in fuel cell electric vehicles

This study introduces a hierarchical model predictive control (HMPC) algorithm utilizing multihorizon optimization for the energy management strategy (EMS) in connected fuel cell electric vehicles (FCEVs). The algorithm leverages the benefits of both long horizon and short horizon decision-making, can solve problems globally, and ensures high computational efficiency. By combining the multihorizon optimization method, the upper layer plans the real-time reference state of charge (SOC) trajectory of the battery efficiently over the long shrinking horizon. In the lower layer, real-time energy optimization is performed with specific objectives such as hydrogen consumption, component lifespan, and following the reference SOC. Additionally, this paper focuses on a nonlinear MPC solution algorithm based on control variable parameterization (CVP). By parameterizing control variables, the number of control sequences to be solved within the prediction horizon is reduced. The algorithm demonstrates significant advantages in rapid solving. Through evaluation in hardware-in-the-loop (HIL) experiments, compared with the traditional MPC algorithm, the proposed algorithm exhibits cost savings ranging from 4.46% to 5.51%, and the computational efficiency of the real-time optimization is improved by 19.94%.