基于MPC的多车纵向协同避撞算法
MPC based longitudinal coordinated collision avoidance for multiple connected vehicles

车辆防撞预警系统在近年得到快速发展，有效提升了道路交通的安全性。然而现有避撞算法大多考虑相邻两车间的安全关系，并未涉及前车行为对连续后车造成的不利影响。该文提出一种基于模型预测控制（model predictive control，MPC）的多车纵向协同避撞算法，通过综合考虑队列运动学关系，基于模型预测控制，以总相对动能密度最小作为优化目标，将连续时间优化问题离散化为二次规划的数值计算，以避免或减少队列整体的碰撞损失。最后通过MATLAB进行数值仿真，验证了所提方法的有效性。
Abstract：The rapid development of vehicle collision warning systems has improved the road traffic safety. Current collision avoidance strategies, however, solely focus on the relationship between two consecutive vehicles, while ignoring the influence of avoidance strategies on the traffic flow behind the vehicle. A longitudinal coordinated collision avoidance (CCA) strategy was developed based on model predictive control (MPC) for multiple connected vehicles. The model considers multiple vehicle flow dynamics to minimize the total relative kinetic energy density (RKED) as the objective with and the optimization problem in continuous time discretized into a quadratic programming problem for numerical calculations to avoid or reduce the total loss of vehicle collisions. The effectiveness of this method is verified by MATLAB simulations.