Longitudinal maneuver design in coordination layer for automated highway system

The automated highway system (AHS) architecture of the California PATH program organizes traffic into platoons of closely spaced vehicles and has a layered control system. Maneuvers in AHS include join, split, lane change and merging/exit to and from the automated lanes. The main function of the coordination layer is to coordinate all the maneuvers of each vehicle in a specified section of AHS such that they are performed in safe, smooth and deterministic manner. Coordination layer designs maneuver by determining maneuver parameters and calculate reference trajectory for each vehicle, which keeps the feasibility of maneuvers and guarantees the safety of passengers, ride quality, and avoiding control saturation. Furthermore, it is desired that the decisions are made in an optimal manner to minimize time and/or other performance criteria like fuel consumption while constraints on variables are not violated. This paper presents the mathematical model and design procedure for coordination layer control using a hybrid system approach which unifies coordination layer and regulation layer. It is shown that this problem can be formulated as a mathematical programming problem, which can be easily modeled using AMPL modeling language software and readily solved by the MINOS solver.