摘要
城市轨道交通站点的精准短时客流预测可以很好地缓解城市交通拥堵,给城市居民带来更快速、更优质的出行服务。通常短时客流预测随时间的变动而变动,长短时记忆网络能对其进行深度的训练和特点提取。为提升预测性能,以成都轨道交通火车北站为例,设计一种基于改进PSO-LSTM模型的城市轨道交通站点短时客流预测办法。通过实例研究分析,验证了改进后的PSO-LSTM模型在城市轨道交通站点短时客流预测中具有更好的预测性能。
Accurate short-term passenger flow prediction of urban rail transit stations can well alleviate urban traffic congestion and bring faster and better travel services to urban residents.Generally,the short-term passenger flow forecast changes with time,which can be deeply trained and extracted by short and long time memory network.In order to improve the prediction performance,this paper takes Chengdu rail transit north railway station as an example and proposes a short-term passenger flow prediction method of urban rail transit based on the improved PSO-LSTM model.Through case study analysis,it is verified that the improved PSO-LSTM model has better prediction performance in short-term passenger flow prediction of urban rail transit stations.
作者
张国赟
金辉
Zhang Guoyun;Jin Hui(School of Automobile and Traffic Engineering,Liaoning University of Technology,Jinzhou 121001,Liaoning,China)
出处
《计算机应用与软件》
北大核心
2021年第12期110-114,134,共6页
Computer Applications and Software
关键词
城市轨道交通
站点短时客流预测
长短时记忆网络
粒子群算法
Urban rail transit
Short-term passenger flow forecast
Long and short time memory network
Particle swarm optimization
作者简介
张国赟,硕士生,主研领域:智能运输系统与物流配送;金辉,教授。
