Determination of early green time for e-bikes at signalized intersections

Based on the characteristics of interference mechanism between motor vehicles and non-motorized vehicles, the paper proposes a method for determining early green time of e-bike and constructs a multi-objective optimization model of mixed traffic flow signal timing. Combined with video extraction technology, the floor map is used to analyze the interaction between e-bikes and motor vehicles, and the effect strength index is proposed which can reflect interference degree between motor vehicles and non-motorized vehicles. By fitting field data, the composite relationship among initial release time, effect strength and e-bike ratio is determined. Then the early green time models of e-bikes in through and left directions are obtained respectively. Taking into account the time benefit, road use rate and environmental benefit, a multi-objective optimization model for mixed traffic flow signal timing considering e-bike early green is put forward, which it is solved by NSGA-II algorithm. The validity and practicability of proposed models are verified by sensitivity analysis and comparison analysis of different timing algorithms. The results show that, compared with five types of signal timing plans of intersection under different traffic conditions, the timing plan proposed in this paper has obvious advantages. The maximum fluctuations of cycle length, average vehicle delay, capacity and average stops per vehicle are -11.64%, -26.33%, -15.63% and 6.48%. © 2020, Jilin University Press. All right reserved.