关键词: 协同自适应巡航控制系统, 混合流, 交叉口, 借道左转, 元胞自动机

Abstract: In order to improve the traffic efficiency of left-turn traffic flow, some Chinese cities, such as Handan, Shenzhen, Jinan and Changsha, have adopted a non-traditional way of organizing left-turn traffic flow at some intersections: contraflow left-turn lane. The basic idea of contraflow left-turn lane is to set a section of the adjacent opposite lane close to the stop line as a left-turning reverse variable lane at the intersection with a left-turning lane and a left-turning phase. Under certain signal and safety conditions, left-turning vehicles can enter the lane in advance and wait for release. As a traffic organization mode that can effectively improve the traffic efficiency of left-turning traffic flow, in the face of increasing traffic pressure, the contraflow left-turn lane will be adopted by more and more intersections. In addition, the Cooperative Adaptive Cruise Control system (CACC) developed to solve a series of traffic problems has been in the experimental stage. At present and for a long time to come, there is and will be a mix of conventional manually driven vehicles (MV) and CACC equipped vehicles (CV) on the road.
In this context, in order to provide some theoretical reference for the promotion process of the contraflow left-turn lane, this paper establishes a cellular automata model for the mixed flow of CV and MV under the traffic organization mode, and studies the operation characteristics of the mixed flow in the contraflow left-turn lane scenario. The mature MCD model is adopted for the movement rules of MV. The movement rules of CV are adapted from the MCD model according to their characteristics different from MV (they are always sensitive to external dynamics, can communicate with each other, and adjust the following strategy according to the type of vehicle ahead, etc.). The key point of the model is to solve the problem of whether CV enter the contraflow left-turn lane when driving to the open area. The decision to enter the contraflow left-turn lane depends on whether delays are reduced so that vehicles can pass earlier. The challenge is to compile an algorithm for the CACC so that it can estimate as accurately as possible the passage time in both entering and non-entering cases. In order to establish rules for vehicles entering the contraflow left-turn lane, we consider location conditions, signal conditions, safety conditions and vehicle types. CV decide whether to advise drivers to enter the opposite lane based on the calculated passage time for both conditions. The passage time is affected by the number and state of vehicles ahead. The situation ahead can be divided into three categories: no vehicles, only stationary queuing vehicles and moving vehicles. These three categories require different passage time calculation methods.
We have written the model into MATLAB software for a lot of simulation and analysis, with the purpose of studying the impact of market penetration of CV on the left-turn traffic. The results show that, with a gradual increase in CV market penetration, the total passing volume of the three left-turn lanes increases slowly, and the passing volume of pure CACC vehicles increases by 18.5% compared with that of pure MV vehicles, indicating that the CACC can improve the vehicle passing efficiency and increase the flow. The passing volume of the two traditional left-turn lanes increases gradually with the market penetration of CV. The left traditional left-turn lane has slightly less traffic than the conventional lane on the right, because U-turn vehicles have to wait in the left lane for the right time to U-turn. The passing volume of the contraflow left-turn lane gradually decreases, that is to say, fewer and fewer vehicles need to enter this lane. It should be due to the improvement of the efficiency of the traditional lane and the reduction in the demand for vehicles to enter the contraflow left-turn lane. Moreover, with an increase in CV market penetration, the traffic volume of U-turn vehicle increases gradually. In short, the CACC system can improve the traffic efficiency of left-turn vehicles.

Key words: cooperative adaptive cruise control system, mixed traffic flow, intersection, the contraflow left-turn lane, cellular automata model