Coordinated Vehicle Platooning with Fixed Routes : Adaptive Time Discretization, Strengthened Formulations and Approximation Algorithms

We consider the coordinated vehicle platooning problem over a road network with time constraints and the routes of vehicles are given. The problem is to coordinate the departure time of each vehicle to enable platoon formation hence maximizing the total fuel saving. We first focus on the case that the routes form a tree. In this case, the flexibility time window of each vehicle admits a unified representation, under which the synchronization of departure time is equivalent to the management of time-window overlapping. Based on this observation, a time-window-overlapping induced mixed-integer linear program formulation has been established which is equivalent to the continuous-time formulation established in the previous work. By imposing an adaptive time-discretization procedure, we can further reformulate the time coordination problem using a vehicle-to-time-bucket assignment induced mixed-integer linear program. Compared to the continuous-time formulation, the assignment formulation is free of big-M coefficients, and it is demonstrated by numerical experiments that the assignment formulation is indeed more effective in computational performance.As an independent interest, three approximation algorithms have been derived with provable competitive ratios under a certain regularity assumptions, which is for the first time on this discrete optimization problem. The solution approach has been extended to general cases in which the graph formed by vehicle routes can have loops, by adding a systematic loop breaking scheme. An insightful discovery is that whether there exists an exact equivalent assignment formulation for a general problem instance has a close connection to the pattern of lattice groups. A systematic numerical investigation has been conducted to demonstrate the effectiveness of the assignment formulation, the adaptive time discretization and the approximation algorithms