Optimization of Tractor Battery Temperature Control Performance Based on Piecewise Linear Model Predictive Control

To address the challenges of high thermal loads and limited energy efficiency in an electric tractor operating under complex agricultural conditions, this paper proposes a hierarchical battery thermal management strategy based on liquid cooling. The method integrates an upper-level piecewise linear model predictive control to regulate battery temperature and a lower-level convex optimization scheme for dynamic actuator power allocation among the compressor, cooling fan, and expansion valve. By decomposing the nonlinear thermal dynamics into multiple local subregions, the predictive accuracy is enhanced while maintaining real-time computational feasibility. Comparative simulations reveal that under severe 45 °C ambient conditions, the proposed strategy limits the maximum temperature difference among battery cells to 1.34 °C and average temperature fluctuations to 0.231 °C, significantly outperforming conventional linear baseline methods which resulted in 1.66 °C and 0.349 °C, respectively. Furthermore, the optimized actuator coordination reduces total cooling energy expenditure by 11.4%, effectively minimizing transient peak loads on the high-voltage bus and preserving energy for primary traction tasks. These quantitative results confirm that the proposed control framework substantially improves battery thermal stability and powertrain energy efficiency, demonstrating robust potential for practical implementation in heavy-duty agricultural machinery.