A Hierarchical Planning and Predictive Strategy for Integrated Thermal Management in Electric Vehicles: Optimizing Energy Efficiency and Cabin Climate

Abstract

Thermal management systems (TMS) in electric vehicles (EVs) face significant challenges in balancing energy efficiency, battery safety, and cabin comfort, particularly under extreme weather conditions and dynamic driving scenarios. Traditional real-time optimization strategies, such as model predictive control (MPC), often fail to address long-term energy sustainability, while global optimization methods like dynamic programming (DP) are computationally intensive and impractical for real-time applications. To address these limitations, this study proposes a novel hierarchical DP-MPC framework that integrates the global optimization capabilities of DP with the real-time adaptability of MPC. The DP layer optimizes long-term energy allocation and temperature trajectories, while the MPC layer fine-tunes local control actions to adapt to real-time changes. Simulation results demonstrate that the DP-MPC framework significantly reduces energy consumption, maintains battery temperature within the optimal range, and ensures cabin comfort, outperforming traditional PID and standalone MPC strategies. The proposed framework also exhibits strong robustness in handling uncertainties such as fluctuating ambient temperatures and traffic conditions, offering a scalable and computationally efficient solution for real-time thermal management in EVs. This research provides a foundation for the development of more intelligent and adaptive thermal management systems, contributing to the broader adoption of electric vehicles and the transition to sustainable transportation.

Keywords: Electric vehicles, thermal management, dynamic programming, model predictive control, Energy efficiency, cabin comfort