Synthesis of a behavior-guided controller for lead vehicles in automated vehicle convoys

Reliable cooperative adaptive cruise control on highways requires the lead vehicle of a vehicle convoy to be capable of resisting disturbances outside the convoy. This paper proposes a controller synthesis approach adopting behavior classification to improve the lead vehicle's ability to deal with outer disturbances. First, a behavior classifier is developed based on hidden Markov models to detect dangerous driver behaviors of surrounding traffic participants. The classification results with corresponding predicted trajectories are then imported to a model predictive controller for the lead vehicle. A behavior-guided cost function of the controller is carefully designed to react to behavior differences and to contribute to convoy string stability. The impact of the lead vehicle's state deviation on the convoy is studied based on leader-to-formation stability properties. Furthermore, a nonlinear bound is also given to state the performance of the proposed controller. Simulations of a cut-in scenario are conducted using the CarSim simulation environment to show the effectiveness of the proposed controller.