摘要
【背景】由于轨道交通线网的不断完善,城市地铁站点数量也随之增多,轨道交通公司需要对数百个客流特征不同的站点训练各自的客流预测模型。传统方法是对站点逐个进行训练,耗费时间长,无法满足高时效性的要求。【目标】通过研究客流预测模型迁移学习方法,仅对少数站点进行客流预测模型训练,其余站点进行模型迁移学习训练,提升全线网站点的预测效率。【方法】考虑波动特征相似度的预测模型迁移学习,通过提取每个站点的客流波动特征,计算当前站点与其他站点的波动特征相似度,为每个站点匹配与其相似度最高的站点作为迁移学习对象站点,实现“一站一方案”。该方法仅需对少数站点进行正常的预测模型训练,其他站点都可以使用迁移学习的方式训练各自的预测模型。【数据】以成都市地铁为例,选取1、2、3、4号线路上部分站点2022年5月到8月的进出站客流数据进行案例分析。【结论】选取四类不同的深度学习客流预测模型,在使用迁移学习方法进行训练后有效地减少了各类模型的训练时间,所选取站点的训练时间平均减少62.68%,部分站点训练时间能减少80%以上,并且在减少训练时间基础上,四类预测模型的预测精度仍有小幅提升,验证了该方法具有可行性。
[Background]The continued expansion and optimization of urban rail transit networks has led to a marked growth in the number of metro stations has increased substantially.This growth demands the development of individualized passenger flow prediction models for hundreds of stations,each with distinct ridership patterns.Conventional station-by-station training methods are timeconsuming and fail to meet the stringent temporal efficiency requirements of contemporary transit management systems.[Objective]To address this issue,we explored transfer learning techniques for passenger flow prediction models,training them exclusively on select prototype stations.Subsequently,the learned parameters are transferred to the remaining stations through transfer learning,enhancing the prediction efficiency throughout the network.[Method]We propose a transfer learning-based prediction model that considers similarities in passenger flow fluctuation characteristics.By extracting each station's passenger flow fluctuation characteristics,we compute similarity scores with other stations and assign each to the most similar station as its transfer learning target,achieving a"one station,one solution"strategy.This method requires extensive prediction model training for only a few stations while allowing others to leverage transfer learning.[Data]For analysis,the Chengdu Metro served as a case study,utilizing passenger flow data of selected stations on Lines 1,2,3,and 4 from June to August 2022.[Conclusion]We evaluated four deep-learning passenger flow prediction models.Transfer learning reduced training time by an average of 62.68%,with some stations experiencing over 80%reduction.Despite the decrease in training duration,prediction accuracy slightly improved across all models,confirming the efficacy of this method.
作者
黄嘉
赵玲
HUANG Jia;ZHAO Ling(Chengdu Metro Operation Co.,Ltd.,Chengdu 610051,China;Chengdu Zhiyuanhui Information Technology Co.,Ltd.,Chengdu 610213,China)
出处
《交通运输工程与信息学报》
2026年第1期50-63,共14页
Journal of Transportation Engineering and Information
基金
四川省科技计划项目(2023ZHCG0018)。
关键词
城市交通
客流预测
迁移学习
地铁客流
客流波动特征
urban transportation
passenger flow forecast
transfer learning
subway passenger flow
passenger flow fluctuation characteristics
