Realistic characteristics and asymmetric model of heterogeneous traffic flow on two-lane freeways

On two-lane freeways with high ratios of fast and slow vehicles, traffic congestion can frequently occur when the traffic volume is significantly lower than the design capacity. This paper studies the traffic flow characteristics on two-lane freeways in the situation. We investigate and analyze detailed traffic flow data collected using roadside laser instrument from a typical two-lane freeway in China. The field data shows that cars and heavy trucks, the main vehicle types, have a huge speed difference. Specifically, in low-density traffic with such high truck ratios that cars cannot freely overtake slow trucks, the traffic flow presents some unique characteristics. A great deal of lane changing occurs during the driving of cars and trucks. Moving bottlenecks occur randomly and frequently in space and time, so the traffic flow state is always unstable. Based on those findings, the active and inactive lane-changing cellular automata model is proposed, and asymmetric lane-changing rules considering cars and trucks are established. The simulated results show that the proposed model can reflect the key traffic flow characteristics. Vehicle segregations, which present vehicles of different types run in different lanes and that vehicles of the same type run in groups, are confirmed by the field data and reproduced by the proposed model. These results can be applied to the existing speed and lane management system of two-lane freeways in China to alleviate congestion and improve capacity.