基于土地利用的城轨新站点客流预测方法研究

城市轨道交通线路规划、设计和建设全过程年限较长,在此期间土地利用规划通常多有调整,导致工可研阶段的客流预测结果难以支撑新线开通时运输计划编制等运营管理工作。提出一种基于土地利用的城轨新站点客流预测方法,首先研究城轨站点周边人口岗位数据统计方法,其次提出人口岗位比、人口岗位总数和站点可达性指标概念,基于K近邻非参数回归预测方法构建进出站量预测模型,最后依托广州地铁客流数据对模型进行精度分析。结果表明,所有站点进站量预测的平均绝对误差占平均实际进站量的19.0%,进站量大于2万人次的站点平均相对误差为16.0%,所提方法可为城轨新线开通后运输组织提供决策依据。

Passenger Flow Forecast of New Urban Rail Transit Stations Based on Land Use

Urban land use planning is adjusted frequently during the continuing overall process of urban rail transit planning, design, and construction. As a result, the passenger flow forecast of the engineering feasibility study stage hardly supports the operation management work when new lines are put into operation. A method based on land use is proposed in order to achieve the passenger flow forecast of new urban rail transit stations. First, a statistical approach to population and posts around urban rail transit stations is proposed. Second, the ratio of population to post, total number of population and posts, and station accessibility indices are introduced to build the passenger flow forecast model using the K-nearest-neighbors (KNN) algorithm. Finally, the entrance passenger flow data of Guangzhou Metro are used for error analysis. The result shows that the mean absolute forecasting error is approximately 19.0% of the mean actual entrance passenger flow for all stations, and the mean relative forecasting error is approximately 16.0% when the station entrance passenger flow is greater than 20, 000, which demonstrates that the proposed method can provide foundation data for transport organizations after new urban rail transit lines operate.