铁道科学与工程学报2025，Vol.22Issue(8)：3436-3447,12.DOI:10.19713/j.cnki.43-1423/u.T20241833
基于有序聚类的分时段停站时分标尺优化
Dwell time scale optimization based on the ordered clustering method
摘要
Abstract
To achieve refined management of train dwell time for urban rail transit system and improve the overall transportation capacity and service level,this paper proposed a refined management mechanism for train dwell time based on the analysis of passenger boarding and alighting demands.By focusing on passenger demands,this mechanism dynamically adjusted the train dwell time to ensure the efficiency,safety,and sustainability of the urban rail transit system.To ensure the universal applicability of the proposed management mechanism across different stations,this paper focused on the peak door of the station,which determined the passenger boarding and alighting time of the train.First,based on the on-site survey data,a segmental formula for passenger boarding and alighting time calculation was established by using a piecewise linear fitting model,which considered the total number of boarding and alighting passengers,train load factors,and peak door coefficients.This formula was then combined with train equipment and driver operation time to gain the train dwell time demands.In recognition of the practical implementation challenges faced by overly flexible train dwell time demands in actual train timetabling,this paper adopted an improved ordered sample clustering technique to design a refined calculation method for train dwell time scales,considering practical constraints such as minimum time segment duration and differentiation.The optimal number of time segments for dwell time scales under different conditions was determined using the time difference method and elbow method,ensuring to finely identify passenger demands and fully implement them in practice.Finally,the Shanghai Metro Line 13 was used as an empirical case to verify the effectiveness of the optimization algorithm.The results show that,compared to the current train dwell time scale,the optimized dwell time scale can better identify changes in train dwell time demands at the station level,with an average error reduction of about 2 seconds.For train timetabling,the optimized dwell times can better conserve transport resources while ensuring safety,with an average train travel time savings of at least 1 minute and an increase in transport resources during peak times by approximately 2%.The research results provide theoretical basis and technical support for the refined management of train dwell time scales in urban rail transit.关键词
城市轨道交通/停站时分/乘客乘降/有序样本聚类/最优划分方法Key words
urban rail transit/dwell time/passenger boarding and alighting/ordered sample clustering/optimal partitioning method分类
交通工程引用本文复制引用
张琦,葛健豪,陈蕴祺..基于有序聚类的分时段停站时分标尺优化[J].铁道科学与工程学报,2025,22(8):3436-3447,12.基金项目
国家自然科学基金资助项目(72101184,72071147) （72101184,72071147）
上海市自然科学基金资助项目(23ZR1467400) （23ZR1467400）
