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

TL;DR

This paper introduces a multi-agent reinforcement learning approach to actively reduce community fear during disasters by controlling communication, power, and emergency management systems, improving resilience based on hurricane data.

Contribution

It extends existing CPS models with control mechanisms and formulates a multi-agent differential game solved via actor-critic RL, demonstrating effective fear reduction.

Findings

Achieves 70% fear reduction in Hurricane Harvey simulations.

Attains 50% fear reduction in Hurricane Irma case, showing model generalizability.

Improves infrastructure recovery alongside fear mitigation.

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.