基于线性路径跟踪控制的换道避撞控制策略研究

为实现车辆自主避撞，改善道路交通安全状况，提出一种基于线性路径跟踪控制的换道避撞控制策略。为实时确定制动和换道时机，获取跟车状态下自车和前车车速、加速度、相对距离以及驾驶人制动反应时间计算制动安全距离和换道安全距离，并在此基础上分别引入制动危险系数B和换道危险系数S评估制动与换道风险，使得车辆发生追尾碰撞的危险程度和主动干预阈值更直观。根据车辆期望横向加速度和期望横向位移的变化特性，采用5次多项式法规划符合驾驶人换道避撞特性的避撞路径。为保证换道避撞过程中驾驶人的安全舒适，采用最大横向加速度约束换道避撞轨迹。为实现对换道避撞路径的线性跟踪控制，保证车辆的操纵稳定性和横摆稳定性，基于车辆稳态动力学模型建立前馈控制，结合线性反馈控制消除换道路径的位置和横摆角偏差，修正参考路径实现直车道场景追尾避撞控制。仿真和实车交叉验证试验表明：根据车辆期望横向加速度和期望横向位移建立的符合驾驶人换道避撞特性的五次多项式换道路径与驾驶人实际换道避撞路径基本吻合，结合碰撞时间和车间时距的制动避撞控制策略能够在保证车辆行驶安全舒适性的同时有效避免车辆追尾碰撞，减少交通事故的发生。

Collision Avoidance by Lane Changing Based on Linear Path-following Control

To avoid the collision of autonomous vehicles and improve the road traffic safety, a rear-end collision avoidance strategy by lane changing based on linear path-following control is proposed. In this system, the speed, acceleration, relative distance between the subject and lead vehicles under car-following condition, and the driver's reaction time were obtained to calculate the safe braking and safe lane-changing distances, to determine the time required for the respective actions. On this basis, the braking hazard coefficient B and the lane-change risk coefficient S were introduced to evaluate the risk in braking and lane changing, respectively. This makes the risk degree of rear-end collision and the threshold of active intervention more intuitive. According to the variation in the expected lateral acceleration and lateral displacement of the vehicle, the 5 times polynomial method was used to plan the collision avoidance path, adaptive to the driver's lane-changing behavior. To ensure the safety and comfort of the driver during the lane-changing collision avoidance process, the maximum lateral acceleration of the trajectory was constrained. Linear tracking control was adopted to avoid collision and ensure the stability of the vehicle. To achieve this, the feed-forward control strategy was established based on the steady-state dynamic model of the vehicle, and the position and yaw angle deviations of the lane-changing path were eliminated by combining the linear feedback and feed-forward controls. Correct the reference path to realize the collision avoidance control of the straight lane scene. The experiments by simulation and real vehicle cross-validation demonstrate that the fifth-order polynomial lane-changing path based on the driver's desired lateral acceleration and the expected lateral displacement is in good agreement with the driver's actual collision avoidance path. The braking and collision avoidance strategy based on reaction time and workshop time interval can effectively avoid vehicle rear-end collisions and reduce the occurrence of traffic accidents, while ensuring the safety and comfort of the passengers.