Alleviating Community Fear in Disasters via Multi-Agent Actor-Critic Reinforcement Learning

2026-04-09

AI Summary

This study addresses the challenge of cascading failures across power, communication, and social behavior during disasters, which amplify community panic and undermine collaborative response efforts—limitations exacerbated by the absence of proactive intervention mechanisms in existing models. To overcome this, the authors propose an extended cyber-physical-social resilience framework incorporating control channels from three institutional actors, formulated as a three-player non-zero-sum differential game. The solution leverages online multi-agent actor-critic reinforcement learning to enable, for the first time, coordinated proactive regulation across interdependent infrastructures and human behaviors. Simulations based on real-world data from Hurricane Harvey demonstrate that the approach reduces community panic by 70% on average and significantly accelerates infrastructure recovery. Notably, without retraining, it achieves a 50% reduction in panic under Hurricane Irma scenarios, highlighting strong cross-disaster generalization capability.

Technology Category

• Multiagent Systems: Adversarial Agents
• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Search and Optimization: Learning to Search

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

During disasters, cascading failures across power grids, communication networks, and social behavior amplify community fear and undermine cooperation. Existing cyber-physical-social (CPS) models simulate these coupled dynamics but lack mechanisms for active intervention. We extend the CPS resilience model of Valinejad and Mili (2023) with control channels for three agencies, communication, power, and emergency management, and formulate the resulting system as a three-player non-zero-sum differential game solved via online actor-critic reinforcement learning. Simulations based on Hurricane Harvey data show 70% mean fear reduction with improved infrastructure recovery; cross-validation in the case of Hurricane Irma (without refitting) achieves 50% fear reduction, confirming generalizability.

Problem

• community fear
• cascading failures
• cyber-physical-social systems
• disaster resilience
• infrastructure recovery

Innovation

• multi-agent actor-critic reinforcement learning
• cyber-physical-social resilience
• non-zero-sum differential game
• active intervention
• disaster fear mitigation