针对自动紧急制动（autonomous emergency braking，AEB）系统弯道适应性差及舒适性不佳的问题，提出了一种基于高斯过程运动预测，考虑变曲率弯道和制动舒适性的AEB系统控制策略。基于三次样条曲线建立行车道路模型，对前车进行定位，并计算相对曲线距离。考虑车辆运动的非线性特性以及时间效应，建立基于高斯过程理论的车辆运动预测模型，设计了基于预测碰撞时间的分级预警与制动控制策略。联合仿真结果表明：提出的控制策略能够有效实现车辆的避撞，解决了AEB系统在复杂动态工况下的弯道适应性和制动舒适性问题。

To address the issues of poor corner adaptability and discomfort in the autonomous emergency braking (AEB) system during cornering, a control strategy for the AEB system is proposed. This strategy is based on Gaussian process motion prediction, with variable curvature corners and braking comfort taken into account. A road model is established by using cubic spline curves to locate the obstacle and calculate the relative distance. Taking into account the nonlinear characteristics of vehicle motion and the effect of time, a vehicle motion prediction model based on Gaussian process theory is developed. A hierarchical early warning and braking control strategy based on predicted collision time is designed. The results of the co-simulation show that the proposed control strategy can effectively achieve collision avoidance, solving the AEB system’s curve adaptability and braking comfort problems under complex dynamic conditions.

U461.91

国家自然科学基金资助项目（51875061）。