Neural network surrogate models for aerodynamic analysis in truck platoons: Implications on autonomous freight delivery

Abstract

Recent advances in connected vehicles have the potential to revolutionize the efficiency and sustainability of transportation. In particular, truck platooning has emerged as a promising solution for improving freight delivery operations. However, the generalization of truck platoon modeling and the economic implications of truck platoons require further investigation. In this paper, we proposed a data-driven neural network surrogate model to predict the drag force of the truck platoon system. The proposed surrogate model can be generalized to truck platoons of various configurations and allows for the evaluation of fuel consumption reduction of truck platoons. Through a case study on a 100-mile corridor on Illinois I-57 Highway, we demonstrate the substantial fuel savings of up to 10% by truck platooning. Additionally, we conduct a cost-benefit analysis for implementing connected freight delivery systems and highlight the potential for significant reductions in delivery costs per parcel, up to 26%. These findings contribute valuable insights into optimizing truck platooning configurations, showcasing the potential benefits of connected freight operations, and improving environmental sustainability.