Problem
Research questions and friction points this paper is trying to address.
evacuation behavior modeling
heterogeneous behavior
moving threat
crowd simulation
public plaza
An LLM-powered Agent Framework for Heterogeneous Evacuation Behavior Modeling under a Moving Threat in a Public Plaza
🤖 AI Summary
This study addresses the limitation of traditional fixed-rule models in capturing internal human decision-making processes, such as perception and memory, under mobile threats. To this end, it proposes a large language model (LLM)-driven agent evacuation framework. The approach integrates private symbolic views, memory graphs, and persona-based prompting to simulate individual cognition. Furthermore, it introduces a novel paradigm comprising state-compressed context and a validation engine that decouples behavioral selection from physical feasibility, enabling auditable, end-to-end heterogeneous behavior generation. Experimental results demonstrate a strong correlation between knowledge of available exits and evacuation success rates (89.5% vs. 1.05%), revealing the critical influence of personality traits and information acquisition on hazard assessment and evacuation outcomes.
📝 Abstract
Modeling heterogeneous evacuation behavior under a moving threat is difficult because human perception, memory, and evidence evaluation are not well captured by fixed rules. We propose a novel LLM-powered agent-based framework to represent these internal decision processes. Each pedestrian agent perceives a private symbolic ASCII view, maintains a Memory-based Knowledge Graph derived solely from individual observations, and makes decisions through persona-conditioned prompts under a common sampling configuration. A compressed decision context with stateless memory preserves trial-and-error experience across turns while excluding reasoning traces, and a validation engine separates behavioral choice from physical feasibility by executing routes only over observed terrain. We evaluated eight personality compositions in eight paired randomized blocks within a simulated public plaza. Usable-exit knowledge was strongly associated with evacuation success: 89.5% of agents possessing such knowledge evacuated, compared with 1.05% of those without it. Personality compositions also differed in their evaluation of remembered threat evidence: the proportion of danger assessments varied by 0.265, while high-urgency, low-directness decisions ranged from 11.41% to 34.33%. After direct threat sightings, responses converged, with 99.6% of assessments classifying the situation as dangerous. Movement was selected in 99.8% of decisions. Overall, evacuation outcomes were strongly associated with information access, while evacuation time was jointly associated with spatial geometry, information, and affect. The framework provides an auditable approach to generating endogenous behavioral heterogeneity through persona-conditioned LLM agents in crowd-evacuation simulations.
Innovation
Methods, ideas, or system contributions that make the work stand out.
LLM-powered Agent
Heterogeneous Evacuation Behavior
Memory-based Knowledge Graph
Persona-conditioned Prompts
Validation Engine
🔎 Similar Papers
No similar papers found.
