城市轨道交通非常态客流出行影响机制研究

Influence Mechanisms of Abnormal Passenger Flow Travel in Tianjin Urban Rail Transit

为探究异常天气情况下多维空间属性对轨道交通非常态客流出行机制的影响,基于梯度提升决策树模型,分析异常天气下天津市143个轨道交通站点的非常态客流出行特征,解析经济属性、交通便利性、建成环境等10个影响因子对非常态客流量的非线性影响机制。研究发现:①异常天气下,轨道交通站点日均客流显著增长。受天气状况的影响,97%的轨道交通站点日均客流呈增长趋势,其中天津站客流增长幅度最大,达到32%;②建成环境对城市轨道交通非常态客流的影响最为显著,达到86.44%,因子“到市中心距离”影响程度最大;③各影响因子与非常态客流之间具有明显的非线性关系与阈值效应,除公交站点数量、交叉口数量以及居住设施密度与非常态客流间呈波动“凸”形非线性关系外,其他影响因子整体保持非线性正相关或负相关。

In order to investigate the influence of multidimensional spatial attributes on the abnormal passenger flow travel mechanism of rail transit under abnormal weather conditions.Based on the gradient boosting decision tree model,we analyzed the extraordinary passenger flow travel characteristics of 143 stations in Tianjin under abnormal weather and resolved the nonlinear influence mechanism of 10 influencing factors such as economic attributes,traffic convenience and built environment on the abnormal passenger flow.The study found that:①Under abnormal weather,the average daily passenger flow of rail transit station increases significantly.97%of the average daily passenger flow at rail stations tended to increase due to weather conditions,with the largest increase of 32%at Tianjin Station.②The built environment has the most significant impact on the abnormal passenger flow of urban rail transit,reaching 86.44%,and the factor to the city center has the greatest degree of influence on the distance.③There is a significant non-linear relationship and threshold effect between each influence factor and the abnormal passenger flow.Except for the number of bus stops,the number of intersections and the density of residential facilities,which show a fluctuating“convex”non-linear relationship with abnormal passenger flow,all other influencing factors maintain a non-linear positive or negative correlation.