智能网联环境下基于安全势场理论的车辆换道模型

为有效刻画未来智能网联环境下车辆在换道过程中面临的驾驶风险，保证车辆执行更加安全的换道决策，建立基于安全势场理论的车辆换道模型。首先针对车辆换道过程中所遇到的驾驶风险进行评估，利用势场理论给出车辆行驶过程中不同运动状态下安全势场的空间分布。其次根据换道过程中相关车辆不同安全势场分布情况计算出换道结束时的车间临界距离，相比于传统的车间临界距离计算模型，提出方法能够动态刻画出车辆在不同速度、加速度条件下临界距离的变化趋势，并且能够根据车辆不同的运动状态，动态表达出车辆间临界距离的变化。在此基础上，根据智能网联环境下车辆各类运动状态能够被实时感知的特点，总结出车辆各类运动状态下需要的换道安全临界时间，最终建立基于安全势场理论的最小安全距离换道模型。最后，对模型进行数值仿真分析，仿真结果表明：车辆换道所需要的最小纵向安全距离与换道车辆以及其周围车辆的运动状态有着直接关系。在今后趋于成熟的智能网联环境下，该模型可以进一步进行扩展，利用安全势场的分布情况，对车辆换道过程进行动态实时干涉，能够为今后智能网联环境下车辆协同换道、车辆自动驾驶以及车辆群体优化控制等相关研究提供一定的理论支撑。

Lane-changing Model Based on Safety Potential Field Theory Under the Connected and Automated Vehicles Environment

To characterize the driving risks that vehicles face during the lane changing process in a future connected and automated vehicles (CAVs) environment effectively and ensure that vehicles make safer lane change decisions, a vehicle lane changing model based on the safe potential field theory was established in this study. First, the driving risk encountered during the vehicle lane changing process was evaluated and the potential distributions of the safety potential field under different motion states during the vehicle driving process were identified based on the potential field theory. Second, the critical distances between vehicles at the end of lane changing were summarized according to the distribution of different safety potential fields for relevant vehicles during the lane changing process. Unlike the traditional distance calculation model for vehicles, the method proposed in this paper can dynamically characterize the trends of the critical distance under different speed and acceleration conditions. According to these characteristics, various types of vehicle movement statuses can be perceived in real time in a CAVs environment. Additionally, the critical times required for lane changing under various motion states can be summarized and a model of the minimum safe distance for lane changing can be established based on the safety potential field theory. Numerical simulation analysis of the proposed model demonstrates that it can characterize the effects of various motion parameters on lane changing results. In a CAVs environment, will be perfected in the future, the proposed model can be further expanded to apply the distributions of the safety potential field to intervene in the vehicle lane changing process dynamically and perform vehicle group optimization control in real time.