信号交叉口车路环境对电动自行车释放膨胀特性的影响分析

电动自行车数量的急剧增长导致其在绿灯释放阶段膨胀特性明显，进而加重了交叉口的机非冲突、降低了车流的通行效率。利用视频轨迹提取技术，通过光流法的表现形式描述直行电动自行车在绿灯期间的膨胀特征，并根据其密度变化、膨胀差异和电动自行车对机动车的影响程度，确定出释放初期为主要研究时段；同时，提出了一种反映电动自行车膨胀变化的新型指标膨胀度，分别通过线性相关分析、秩相关性分析和偏相关分析，确定了车路环境中影响膨胀度的动态因素和静态因素；最后基于6个信号交叉口的实测数据，建立各因素与膨胀度的数学关系模型，并结合实际交通条件，给出不同车路环境下电动自行车的管控措施与渠化方法。研究结果表明：车路环境中的电动自行车流量、机动车流量、电动自行车过街距离、非机动车进/出口道宽度、机非分隔带设置情况这5种因素对膨胀度的影响能力各异，右转机动车流量与膨胀度相关性最高。此外，动态因素与膨胀度之间具有确定的函数关系，存在电动自行车与机动车流量均衡效益最大的优势区域；静态因素的差异会导致电动自行车膨胀形式的变化；膨胀度可与动态、静态因素构建复合函数模型。研究成果可为混合交通流的渠化设计和信号配时提供理论依据和技术支持。

Effect Analysis of Vehicle-road Environment at Signalized Intersection on the Release and Expansion Characteristics of E-bikes

The rapid increase in the number of e-bikes has led to obvious expansion characteristics during the green light phase; these, in turn, increase intersection traffic conflicts and reduce their operational efficiency.This study uses video trajectory extraction technology to describe the expansion characteristics of e-bikes during green lights by a presentation of the optical volume method. According to density change, expansion difference and effect degree of e-bikes on motor vehicles, initial stage is determined as the main study period.At the same time, a new index-expansion degree,which reflects the expansion change of e-bikes,was proposed. A linear correlation analysis, rank correlation analysis, and partial correlation analysis were used to determine the dynamic and static factors affecting the expansion degree in the vehicle-road environment. Finally, based on field data of six signalized intersections, mathematical relationship models between various factors and expansion degrees were established by fitting this data. Control measures and channelization methods of e-bikes under different vehicle-road environments were combined and given with actual traffic conditions.Results show that five factors of e-bike volume, motor vehicle volume, e-bike crossing distance, non-motorized vehicle entrance width, and dividing strip have different effects on the degree of expansion; the correlation between right-turn vehicle volume and the expansion degree is the highest.In addition, there is a certain functional relationship between the dynamic factor and expansion degree and there is an advantageous area where e-bikes and motor vehicles have the most balanced volume efficiency. Furthermore, differences in static factors can cause changes in the form of expansion of e-bikes and the expansion degree can be combined with dynamic and static factors to construct a composite function model.The research results can provide a theoretical basis and technical support for the channelization design and signal timing of mixed traffic flow.