摘要
城市轨道交通车站设计时,大多数城市采用预测的城市高峰小时客流作为设计客流。但由于城市轨道交通车站客流的高峰出现时段与城市高峰小时不完全一致,导致某些车站设计客流偏小。为研究城市高峰小时客流与车站高峰小时客流的差异,通过引入车站高峰客流偏差系数,合理确定车站设计客流。以西安市地铁为例,运用最小二乘支持向量机建立预测车站高峰客流偏差系数的模型,得出训练集拟合优度为0.71,测试集预测平均相对误差为2.41%,模型拟合效果良好,表明最小二乘支持向量机能够很好地预测车站高峰客流偏差系数。
When designing urban rail transit stations,most cities adopt the forecast city peak hour passenger flow as the design passenger flow,but the station peak passenger flow hour does not exactly coincide with the city peak hour,so some stations are designed smaller than needed.In order to study the difference between the city peak hour passenger flow and the station peak hour passenger flow,the concept of station peak passenger flow deviation coefficient is introduced to rationally decide the station design passenger flow.Taking Xi′an Metro as an example,the model of predicting the station peak passenger flow deviation coefficient is established by using the least squares support vector machine.The goodness of fit of trained set is 0.71,and the mean relative error of the test set prediction is 2.41%,and the model fitting effect is good,which means the least squares support vector machine is a good predictor of the station peak deviation coefficient.
作者
余丽洁
肖娜
陈宽民
马超群
YU Lijie;XIAO Na;CHEN Kuanmin;Ma Chaoqun(College of Transportation Engineering,Chang′an University,710064,Xi′an,China)
出处
《城市轨道交通研究》
北大核心
2021年第1期97-100,共4页
Urban Mass Transit
基金
国家自然科学基金项目(71871027)
陕西省自然科学基金项目(2016JM5063)。
关键词
城市轨道交通
车站
车站高峰客流偏差系数
最小二乘支持向量机
urban rail transit
station
station peak passenger flow deviation coefficient
least squares support vector machine
