关键词: 交通工程；智能驾驶车辆；接近区域；优化车辆组；混合路段；安全条件

Abstract: As urban development progresses, the connected and autonomous vehicles (CAVs) are gradually replacing some human-driven vehicles (HDVs) on road segments. However, HDVs will not be completely replaced in the foreseeable future. Therefore, the mixed traffic environment of CAVs and HDVs that has emerged at this time is the driving environment we are currently facing and will face in the future. The driving behavior of CAVs and HDVs interferes with each other on road sections, resulting in low efficiency of hybrid driving. In order to mitigate the interaction between CAVs and HDVs, a phased dynamic lane guidance algorithm for separating CAVs and HDVs in mixed traffic environments (SCHME) was proposed. The proposed algorithm separated the mixed flow vehicle set in the upstream section of the intersection, adjusted the driving routes of CAVs and performed real-time dynamic updates. Under the condition of meeting kinematic constraint and convergence, HDVs adjusted the routes according to the dynamic routes of CAVs, achieving an increase in vehicle operating efficiency in mixed driving environments within the road section under the condition of the minimum generalized safety loss cost for each vehicle. The operating state of vehicle before entering the intersection was simulated by MATLAB. It is concluded that the SCHME algorithm can improve the average vehicle traffic efficiency within a road section by 17% under the condition of the minimum generalized safety loss cost. At the same time, when the vehicle optimization array is larger and the vehicle set is farther away from the intersection, the lower the penetration rate of CAVs, the lower the generalized safety loss cost per vehicle.

Key words: traffic engineering; intelligent connected vehicles; proximity zone; optimized vehicle groups; mixed human-machine driving environment; safety conditions