Deciding Truck Platooning Operational Rules With Stochastic Approximation for Vehicle Safe Overtaking on Two-Lane Undivided Highways

Truck platooning is a promising technology to improve freight transport efficiency. However, this technology has primarily been explored on multi-lane divided highways, neglecting potential applications on two-lane undivided highways prevalent in rural areas. This paper addresses the challenges of deploying truck platooning on such highways, focusing on overtaking safety. We formulate safe overtaking via stochastic approximation as a feasibility problem. An algorithm is proposed to determine operational rules for truck platooning, considering factors including traffic flow, road segment geometry, and vehicle dynamics. Values for truck platooning operational parameters such as speed, inter-truck following gap, and number of trucks in platoon are decided which ensure overall overtaking safety. The algorithm is iterative and depends on first order information, avoiding possible complex computations of stochastic objective with respect to truck platooning operational parameters. The algorithm is general and adapts to different roadway and traffic conditions and weights on platooning parameters. We show that computation of the algorithm is linear in sample size used in approximation. Simulations show that the algorithm-generated platooning rules offer an overtaking safety guarantee of up to 93% on a straight and flat road segment with sparse traffic.