成都大运会期间地铁站大客流预测及仿真优化

Prediction and Simulation-based Optimization of Large Passenger Flow in Subway Stations During the Universiade

2023年即将到来的成都大运会为城市轨道交通大客流组织提出新要求。对此,基于成都市历史大型活动地铁车站客流数据,建立改进的卷积长短期记忆神经网络模型,以成都大运会凤凰山体育公园的活动散场客流为例进行预测与验证,结果表明模型具有较高精度。建立车站仿真模型并输入预测客流模拟,找到车站现有控制方案存在的瓶颈,提出优化措施并进行验证。研究结论为成都大运会城市轨道交通运输组织方案提供参考,并可应用于其他大型活动预案。

The upcoming Universiade in 2023 puts forward new requirements for the organization of large passenger flow in urban rail transit in Chengdu,Sichuan.Based on the historical passenger flow data in the urban transit in Chengdu,an improved convolutional long short-term memory neural network model is established,and the passenger flow of the event in the Phoenix Mountain Sports Park is taken as an example for prediction and verification.The result shows that the model has high accuracy.Establishing a simulation model of the relevant station based on AnyLogic and inputing the predicted passenger flow for simulation,it finds the bottlenecks in the station's control measures,based on which optimization measures are proposed and proved to be efficient.The research conclusion provides a reference for the urban rail transit organization plan during the Universiade,and can be applied to other large-scale event plans.