Fast Maneuver Planning for Cooperative Automated Vehicles

Abstract

A lane following and lane changing maneuver planning method for automated vehicles is investigated, which is capable of evaluating and incorporating cooperative agreements between several automated vehicles. An application level cooperation protocol is discussed, which allows vehicles to negotiate space time reservations in conflict areas via Vehicle-to-Vehicle communication. The planning method is based on decoupling of longitudinal and lateral movement directions, formulation of convex quadratic programming problems and input-output linearization for recovery of a full state reference trajectory and feed forward controls. Several different lane following and merging maneuvers can be planned in one update cycle in order to support an informed selection of the currently best driving strategy. We demonstrate and evaluate the communication protocol and the maneuver planning method on cooperative lane changing scenarios with a physical automated vehicle as well as in a real time simulation.