Chance-Constrained Stochastic MPC With Adaptive Optimization Horizon and Multitimescale for Electric Vehicle Battery Thermal Management
针对电动汽车电池热管理系统，提出一种结合自适应优化时域和多时间尺度的机会约束随机模型预测控制策略，在保证温度安全的同时降低能耗，仿真显示约束违反减少84.88%，能耗降低2.44%。
Battery capacity and safety are closely related to the battery temperature. The battery thermal management (BTM) system consumes considerable energy to maintain the temperature of the battery in the safe range. This energy consumption significantly decreases the driving range of the electric vehicle (EV). This article investigates the optimal control strategy of the BTM system based on model predictive control (MPC) for the connected and automated EV (CAEV), which minimizes energy consumption of the BTM system under the constraint of power and thermal at the same time. The slow thermodynamics of the battery requires a long prediction horizon to achieve optimal temperature and energy consumption of the BTM system. However, long preview information, such as vehicle speed, has large uncertainties, which significantly affects the energy efficiency performance and constraint enforcement robustness. In this study, the effects of the different prediction horizon lengths and the information within the prediction horizon on the MPC performance are first analyzed. Then, the MPC optimization strategy based on adaptive optimization horizon and multitimescale (AOH-MT) is proposed to reduce the temperature constraint violations and computation time. Finally, to improve the robustness under the real driving condition where there are large uncertainties in the speed preview information, a chance-constrained stochastic MPC (C-SMPC) is proposed and the AOH-MT framework is integrated into its prediction horizon to reduce time cost. The simulation results under real-world traffic data show that the proposed approach reduces the constraint violation by 84.88% and the energy cost by 2.44%, which improves robustness against uncertainty in the speed preview information.