A Fast Heuristic for Finding Near-Optimal Groups for Vehicle Platooning in Road Networks

Grouping vehicles into platoons with short distance can reduce fuel consumption up to 21% [1] and improve capacity of roads. The Vehicle Platooning Problem deals with route planing to form platoons and therefore minimize the overall costs of all considered vehicles. This article is focused on the subject to find groups with most fuel savings potential in acceptable time. We propose a novel spatial grouping algorithm to determine near-optimal groups. Our approach uses fast geometrical heuristics, consisting of direction-comparison, a modified version of a geometric-median-calculation and a comparison of intersections areas between two vehicles respectively. For evaluations is same-start unlimited ILP (SSU ILP) used to solves the Vehicle Platooning Problem to get the optimal solution. Driving in found platoons saves round about 2% to 3% fuel in average compared to the sum of particular shortest paths of the vehicles. The algorithm is tested in simulations on randomly created vehicles on different graphs with the size of 5.000, 10.000 40.000 and edges and round about 0.5 times nodes respectively. The performance is evaluated and the results are compared to the total possible amount of savings.