ABSRACT

An autonomous emergency braking (AEB) algorithm for snow-asphalt joint pavement is proposed. Based on machine vision and kinetic analysis, we realize the identification of road information, including the road type, slope, and road section lengths. A new safety model, the reference velocity model, is proposed to solve the problem of determining the braking time on the joint pavement to achieve collision avoidance. In the asphalt section, we design the desired deceleration considering the comfort and safety, while in the snow section, we use the estimated maximum deceleration that the pavement can provide. To meet the desired deceleration requirement, we choose a single-neuron proportion integration differentiation (PID) controller with a Kalman filter. The joint simulation with CarSim and Simulink shows that the host vehicle successfully realizes collision avoidance in various working conditions and verifies the proposed AEB algorithm. Benefitting by the recognition of the forefront road conditions, our proposed model performs better than the traditional AEB model.

摘要

提出了一种适用于雪-沥青联合路面的自动紧急制动（AEB）算法。基于机器视觉和动力学分析，我们实现了道路信息的识别，包括道路类型、坡度、路段长度。提出了一种新的安全模型——参考速度模型来解决接缝路面制动时间的确定问题，以实现防撞。在沥青路段，我们设计了考虑舒适性和安全性的所需减速度，而在雪地路段，我们使用了路面可以提供的估计最大减速度。为了满足所需的减速要求，我们选择了带有卡尔曼滤波器的单神经元比例积分微分 (PID) 控制器。与CarSim和Simulink的联合仿真表明，主车在各种工况下成功实现了防撞，验证了所提出的AEB算法。受益于对前沿路况的识别，我们提出的模型比传统的 AEB 模型表现更好。