The impact of traffic accidents on traffic capacity in weaving area of highway under the intelligent connected vehicle environment

Connected automated vehicles (CAV) will become an important part of future traffic flows. In the highway weaving area, the road conditions are complex and the possibility of traffic accidents is relatively high. In this paper, a cellular automaton model of heterogeneous traffic flow weaving area of multi-lane expressway under accident conditions is proposed, in which three common accident types of rear-end accidents, lateral collisions and collisions with fixed obstacles are considered, and the influence of accidents on traffic flow in intelligent networked environment is simulated and analyzed. The following conclusions are obtained: (1) Collision with fixed obstacle accidents have the least impact on traffic flow in the weaving area, followed by rear-end collisions, and lastly, lateral collision type. (2) CAVs effectively alleviate the traffic congestion under accident conditions and improve the traffic capacity of the weaving area. At the same time, in order to give full play to the characteristics of CAVs, a longer weaving area is needed to improve the traffic capacity of the weaving area. (3) As the length of the weaving area increases, the traffic capacity of the weaving areas generally tends to increase. However, the traffic accidents occurring on the ramp greatly weaken the weaving characteristics of the vehicles in the weaving area, reducing the impact of the length of the weaving area on the traffic flow. Moreover, under the same type of traffic accident, the traffic capacity of traffic accidents occurring on the ramp is significantly higher than that of other lanes.(4) The traffic accidents occurring at the end of the weaving area have a greater negative impact on the traffic flow, especially the type of lateral collision accidents, which greatly diminish the traffic capacity of the weaving area, but the traffic accidents occurring on the ramp do not have such obvious negative effect.