@misc{9234193,
  abstract     = {{Energy management in cold logistics is critical for sustainable cold chain transportation. However, conventional reactive temperature control strategies often induce excessive energy consumption and mechanical wear due to rigid air temperature tracking and inability to anticipate system thermal dynamics. Guided by the Design Science Research (DSR) methodology, this study proposes a simulation-based digital twin and Model Predictive Control (MPC) framework specifically tailored for the cold chain logistics of plant-based foods. First, a 2R2C thermal model was constructed, in which the specific thermodynamic properties of plant-based foods were rigorously mathematically derived to map the dynamic thermal coupling relationship between the carriage and the goods. Crucially, the proposed MPC shifts the control strategy from continuous temperature tracking to strict boundary compliance, treating the product spoilage threshold as a hard constraint rather than an optimization objective.
Through a quantitative, simulation-based comparative analysis, the robustness of the artifact is rigorously evaluated under varying parameters. Results demonstrate that the MPC can achieve significant energy reductions under various conditions, yielding a 34.47% energy reduction in the baseline scenario and a 44.18% reduction under winter conditions. Especially under high packaging and cargo thermal resistance conditions, the baseline PI controller encountered severe thermal decoupling and integral windup, resulting in the refrigeration unit being overloaded and excessive cooling. In contrast, MPC was able to proactively convert the large thermal inertia of the plant-based food into an active energy saving mechanism, significantly extending compressor coasting phases. Furthermore, MPC exhibits superior dynamic robustness in dealing with hardware degradation. As the insulation performance of the wall deteriorated, the energy saving capability compared to conventional control was further enhanced. Ultimately, this study provided a highly robust algorithm architecture for vehicular ECUs, unlocking extreme energy efficiency potential for complex cold chain scenarios.}},
  author       = {{Xu, Xinyu and Li, Junyang}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Energy-Efficient Dynamic Temperature Control for Plant-Based Food Cold Chains: A Model Predictive Control Approach}},
  year         = {{2026}},
}