Towards more sustainable transportation: Receding horizon predictive energy management for powered truck-trailers using probabilistic efficiency models

Heavy-duty trucks are significant contributors to global CO 2 emissions, necessitating innovative decarbonization solutions. Electric-powered trailers that assist the main tractor’s internal combustion engine represent a promising approach. However, their limited energy storage and ’sensor-poor’ modular design, lacking access to the tractor’s internal data, are key challenges that limit their adoption. This paper presents a novel receding-horizon predictive Energy Management Strategy (EMS) that overcomes these challenges. A hierarchical controller that optimizes a flexible ’saving efficiency’ threshold is proposed, rather than a rigid State of Charge (SoC) trajectory commonly used in existing hierarchical EMSs. This ’saving efficiency’ metric quantifies the diesel fuel saved per unit of electrical energy consumed. A Dynamic Programming (DP) algorithm, operating at a high level in a receding-horizon framework, leverages statistical efficiency distributions to determine the optimal threshold. A low-level, real-time controller then activates assistance only when the current estimated efficiency exceeds this threshold. This strategy enables opportunistic, real-time control in a sensor-poor environment. The proposed framework is validated in a comprehensive simulation environment using NREL drive cycles and over 2200 km of recorded real-world test driving profiles. Results demonstrate substantial improvements: in simulation, the proposed RH-EMS increased the net CO 2 savings by 32.3% on NREL drive cycles and by 19% on simulations using real-world driving profiles compared to a standard linear-discharge baseline strategy. The successful development of this EMS represents an important step toward enabling substantial, real-world CO 2 reductions in the heavy-duty logistics sector. • Novel predictive control strategy developed for sensor-poor e-trailers. • Adaptive efficiency threshold replaces rigid State of Charge tracking. • Algorithm validated over 2200 km of real-world heavy-duty driving data. • Simulations using real-world data demonstrate 19% increased CO 2 savings over baseline.