基于客流与空间热力图的轨道交通站域更新潜力预判方法

Predictive Method for Assessing the Renewal Potential of Urban Rail Transit Station Areas Based on Passenger Flow and Spatial Heat Maps

摘要:

在城市存量更新的背景下，将有限的资金和资源投入回报概率大的存量轨道站域，即科学选择具有更新潜力的轨道站域，实现城市空间整合与结构优化显得极其必要。立足于轨道交通客流可持续性发展视角，认为具有更新潜力的轨道站域应具有轨道客流增长潜力及站域活力激发潜力。按照便于捕捉站域客流-热力类型发生变化的目标，采用多源数据，以上海市408个轨道交通车站为例，分别建立2019、2021、2023三个年份可解释机器学习模型，挖掘站点客流的驱动与制约因素，将站域客流?热力互动类型分为4大类8小类，采用空间马尔科夫链方法揭示2019—2021年、2021—2023年站域客流?热力类型转变规律，同时深入探索周围相邻站域客流?热力类型的正向、负向溢出效应，提出客流?热力视角下轨道交通站域更新潜力预判方法，为存量轨道站域城市更新提供科学依据和建议。

Abstract:

Under the background of urban stock renewal， it is essential to strategically allocate limited funds and resources to existing rail transit station areas with higher potential returns. Scientifically identifying stations with renewal potential is critical for promoting urban spatial integration and structural optimization. From the perspective of sustainable development of rail transit passenger flow， this paper argues that station areas with renewal potential should possess both the capacity for passenger flow growth and the potential to stimulate surrounding vitality. To effectively detect changes in passenger flow–heat interaction types， it employs multi-source data， taking 408 metro stations in Shanghai as case studies. Explainable machine learning models are developed for the years 2019， 2021， and 2023 to identify the driving and limiting factors of station passenger flows. Based on this， station area passenger flow–heat interaction types are classified into four main categories and eight subcategories. The spatial Markov chain method is then applied to analyze the transition patterns of these interaction types from 2019 to 2021 and from 2021 to 2023. In addition， it explores the positive and negative spillover effects from neighboring station areas. Moreover， it proposes a method to pre-assess the renewal potential of station areas from the passenger flow-heat interaction perspective， providing a scientific foundation and policy suggestions for the urban renewal of existing rail transit stations.