基于随机森林——隐马尔可夫融合算法的乘客轨迹还原与意图识别

Passenger Trajectory Reconstruction and Intention Recognition Based on Random Forest-Hidden Markov Fusion Algorithm

摘要: 随着城市轨道交通路网的快速扩张，地铁站作为客流集散的关键节点，其设施布局的合理性直接关系到运营效率与乘客满意度。然而，现有的车站设施布局往往依赖于设计经验及行业规范，缺乏对乘客出行意愿的挖掘，易出现动线冲突与高峰时段局部拥堵。针对这一问题，本文提出了一种融合隐马尔可夫模型与随机森林的乘客行为推演模型，构建了考虑乘客出行意愿的动线预测模型，丰富了地铁车站设施布局优化的理论依据，为运营部门进行车站局部改造及客流组织优化提供了可操作的决策支持。

Abstract: With the rapid expansion of urban rail transit networks, subway stations serve as critical hubs for passenger flow distribution. The rationality of their facility layout directly impacts operational efficiency and passenger satisfaction. However, existing station layouts often rely on design experience and industry standards, lacking consideration of passenger travel preferences. This frequently leads to conflicts in movement paths and localized congestion during peak hours. To address this issue, this paper proposes a passenger behavior simulation model integrating Hidden Markov Models (HMMs) and Random Forests. It constructs a movement path prediction model that incorporates passenger emergence preferences, thereby enriching the theoretical basis for optimizing subway station facility layouts. This approach provides actionable decision support for operational departments undertaking localized station renovations and optimizing passenger flow management.