Energy-efficient dynamic temperature control for frozen plant-based food transport using model predictive control

Abstract

Abstract Refrigerated road transport is energy intensive because conventional controllers regulate compartment air temperature reactively while the cargo itself changes temperature slowly. In this work, the potential energy savings from shifting the control reference from air to cargo core temperature are studied using a simulation model of a semi-trailer carrying frozen plant-based food. The controller uses a two-node thermal model to track both the fast air dynamics and the slow cargo thermal response, and a compressor efficiency map to time cooling decisions for periods of higher coefficient of performance. A shrinking prediction horizon matches the remaining transport time, and a two-layer penalty mechanism enforces the frozen-product temperature boundary. Simulation experiments over a 10 h route with door-opening disturbances and ambient temperature variation show that MPC reduces energy consumption from 28.6 to 22.6 kWh, a 21.1% saving, while maintaining the simulated cargo temperature below -18 °C. The saving arises from three control phases: pre-cooling before scheduled door openings, coasting during high-temperature low-efficiency periods, and recovering when ambient conditions improve.