Abstract:With urban residents’ increasing reliance on metro systems for commuting and other daily activities, extreme weather events such as heavy rainfall and flooding impacting the metro system services are becoming increasingly of concern. Plans for such emergency interruptions require a thorough understanding of the potential outcomes on both the system and individual component scales. However, due to the complex dynamics, constraints, and interactions of the elements involved (e.g., disaster, infrastructure, service operation, and travel behavior), there is still no framework that comprehensively evaluates the system performance across different spatiotemporal scales and is flexible enough to handle increasingly detailed travel behavior, transit service, and disaster information data. Built on an agent-based model (ABM) framework, this study adopts a data-driven ABM simulation approach informed by actual metro operation and travel demand data to investigate the impact of flood-induced station closures on travelers as well as the overall system response. A before-after comparison is conducted where the traveler behaviors in disaster scenarios are obtained from a discrete choice model of alternative stations and routes. A case study of the Shanghai Metro is used to demonstrate the ability of the proposed approach in evaluating the impacts of flood-induced station closures on individual traveler behavior under normal operation and a series of water level rise scenarios of up to 5m. It was found that, when the flood-induced station closures only affect a few river-side stations in the city center, the travelers experience only minor disruptions to their trips due to the availability of unaffected stations nearby as a backup. However, as the water level increases and more stations (mainly in the suburban area) are affected, up to 25% of trips are no longer being fulfilled due to the loss of entrances, exits, or transfer links. The system experiences overall less crowdedness in terms of passenger volume and platform waiting time with a few exceptions of increased passenger load due to concentrations of passenger flows to alternative stations under flooding-induced station closures. The proposed approach can be adapted to other disaster scenarios to reveal the disaster impacts on both aggregated and disaggregated levels and guide the design of more spatio- and temporally-targeted emergency plans for metro systems.

Firm-level behavior control after large-scale urban flooding using multi-agent deep reinforcement learning

Modelling the behaviour of corporations during the flood damage recovery process using multi‐agent deep reinforcement learning

Deep reinforcement learning for optimal rescue path planning in uncertain and complex urban pluvial flood scenarios

Multi-agent deep reinforcement learning based decision support model for resilient community post-hazard recovery

Rescue path planning for urban flood: A deep reinforcement learning-based approach

Forecasting Road Network Functionality States During Extreme Rainfall Events to Facilitate Real-Time Emergency Response Planning

Adaptive Reinforcement Learning Model for Simulation of Urban Mobility during Crises

Predicting Human Mobility Flows in Response to Extreme Urban Floods: A Hybrid Deep Learning Model Considering Spatial Heterogeneity

Real-Time Flood Disaster Prediction System by Applying Machine Learning Technique

STFS-urban: Spatio-temporal flood simulation model for urban areas

Optimizing the Post-Disaster Control of Islanded Microgrid: A Multi-Agent Deep Reinforcement Learning Approach.

Evaluating the flooding level impacts on urban metro networks and travel demand: behavioral analyses, agent-based simulation, and large-scale case study

Combined Sewer Overflow and Flooding Mitigation Through a Reliable Real‐Time Control Based on Multi‐Reinforcement Learning and Model Predictive Control

Coordinating Disaster Emergency Response with Heuristic Reinforcement Learning

A deep reinforcement learning model for resilient road network recovery under earthquake or flooding hazards

A Coupled Human and Natural Systems (CHANS) framework integrated with reinforcement learning for urban flood mitigation

Flood mitigation in coastal urban catchments using real-time stormwater infrastructure control and reinforcement learning

Supporting Post-disaster Recovery with Agent-based Modeling in Multilayer Socio-physical Networks

Flood evacuation simulations using cellular automata and multiagent systems -a human-environment relationship perspective

Strategizing Equitable Transit Evacuations: A Data-Driven Reinforcement Learning Approach

Resilience-based post disaster recovery optimization for infrastructure system via Deep Reinforcement Learning