基于梯度提升决策树的城市轨道交通网络运营态势综合评估方法

段力伟1冉松民1陈瑞雪1邝雅静1董少帅1欧光继2

Method of Comprehensive Evaluation on the Operation Situation of Urban Rail Transit Network Based on Gradient Boosting Decision Tree

DUAN LiweiRAN SongminCHEN RuixueKUANG YajingDONG ShaoshuaiOU Guangji
摘要:
以客流规模、运行效率、运营安全以及工作负荷作为一级指标,建立包含20个二级指标的城轨网络运营态势评估指标体系。对运营态势预警等级进行了划分,提出了以梯度提升决策树算法实现网络运营态势综合评估的研究思路。以重庆市轨道交通网络数据为研究对象,进行运营态势预警等级的综合评估验证,验证了网络运营态势综合评估方法的可行性。
Abstracts:
Taking passenger flow scale, operation efficiency, operation safety and workload as primary indicators, an urban rail network operation situation evaluation indicator system comprising 20 secondary indicators is established. Operation situation early warning grades are divided, and the research idea of using Gradient Boosting Decision Tree algorithm to achieve a comprehensive evaluation of operation situation is proposed. Taking Chongqing rail transit network data as research object, by carrying out comprehensive assessment and verification of the operation situation early warning grades, the feasibility of the network operation situation comprehensive evaluation method is confirmed.
论文检索