Predictive cruise control of vehicles with pre-planned acceleration/deceleration command

This work considers the adaptive cruise control (ACC) problem of vehicles and we propose a new ACC algorithm for vehicle cruising systems based on the model predictive control (MPC) technique. In this algorithm, the acceleration/deceleration command of vehicles is pre-planned within the predictive window that the incremental command action is decaying geometrically. Then the command in the predictive window will change along the same direction (i.e. increasing or reducing), which, from the control theory point of view, leads to a smoother command and response of the ACC vehicle. Moreover, the algorithm is to greatly decrease the online computational demand of ACC controllers by reducing the numbers of decision variables of the MPC optimization. In addition, safety is regarded as constraints and car-following requirements are synthesized into the performance criteria of MPC. Finally, several representative traffic scenarios are used to illustrate the effectiveness of the proposed ACC algorithm from comparisons of ride comfort, fuel consumption and safely car-following.