摘要
目前,关于铁路枢纽出租车换乘系统调度方面的探究主要集中在宏观整体优化上,但对调度的智能化和实时性等微观方面研究有所欠缺。首先,提出一种基于改进的目标检测及追踪算法Yolov5-Deepsort和Informer模型的乘客排队时间实时预测方法;其次,基于长短期记忆网络(LSTM)和attention机制对出租车客流进行预测,提出基于乘客需求优先的出租车调度方法;最后,以扬州东站出租车换乘系统为例验证模型的优化效果。结果表明,该方法不仅能有效提升乘客出行体验,且明显降低出租车司机的平均等待时间,对铁路枢纽出租车换乘系统的智能化改进具有指导意义。
At present, scholars in the field of railway hub taxi transfer system scheduling mainly focus on the macroscopic overall optimization, but there is a lack of micro research on intelligent and real-time scheduling. Firstly, a real-time prediction method of passenger queuing time is proposed based on improved target detection and tracking algorithm YoloV5-DeepSort and Informer model. Secondly, the Long short-term Memory(LSTM) and mechanism are used to predict taxi passenger flow, and a taxi scheduling method based on passenger demand first is proposed. Finally, the taxi transfer system of Yangzhou East Railway Station is taken as an example to verify the optimization effect of the model. The result shows that this method can not only effectively improve passengers’ travel experience, but also significantly reduce the average waiting time of taxi drivers, which has guiding significance for the intelligent improvement of taxi transfer system in railway hubs.
作者
张永阳
张文斌
陈堃
李士香
周竹萍
李卫
ZHANG Yongyang;ZHANG Wenbin;CHEN Kun;LI Shixiang;ZHOU Zhuping;LI Wei(Nanjing Zhongshe Intelligent Technology Co.,Ltd.,Nanjing 210013,China;Yangzhou Traffic Construction Management Co.,Ltd.,Yangzhou 225012,China;China Construction Communications Engineering Co.,Ltd.,Beijing 100000,China;School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China)
出处
《交通科技与经济》
2022年第6期53-59,共7页
Technology & Economy in Areas of Communications
基金
国家重点研发计划政府间国际科技创新合作重点专项项目(2019YFE0123800)
中央高校基本科研业务费专项资金项目(30920021142)。
关键词
铁路枢纽
出租车换乘系统
目标检测
排队时间
客流预测
调度优化
railway hub
taxi transfer system
target detection
queuing time
passenger flow forecast
scheduling optimization
