Driver Behavior of E-bikes at the Onset of Green and Yellow at Signalized Intersections in China

摘要

In most Chinese cities, electric bikes and electrically-assisted bicycles, i.e., E-bikes,have drastically increased in recent years and currently comprise the largestproportion of the non-motorized traffic prevailing at signalized intersections.Therefore, proper treatment of E-bikes has become a vitally important issue inimproving operational efficiency and safety performance of signalized intersections.However, fundamental knowledge on the unique operating characteristics and driverbehavior of E-bikes under various conditions is rather insufficient so far.This study aims at statistically analyzing the critical behavioral parameters ofE-bike drivers, and empirically modeling their start-up behavior at the green onsetfollowing a 3-second of red-and-yellow signal and pass/stop decision behavior at theyellow onset following a 3-second of flashing green. Distribution types andparameters of desired speed, start-up time, acceleration rate, perception-reaction time,and deceleration rate were firstly investigated by the use of highly accurate trajectorydata. Then, a temporal-spatial model was developed to interpret the start-up curve andthree binary logistic regression models were built to predict the pass/stop decisionsrespectively for different driver groups. Conclusions supported that the start-up curveof E-bikes could be well described by a quadratic function and the red-and-yellowsignal significantly induced hurry-start. In addition, potential time to stop-line at thedecision point was found to be the dominant independent factor to explain thepass/stop decision of E-bike drivers, and the flashing green signal seemed to enlargethe option zone, bring the indecision zone earlier, and result in more aggressive pass.