Parallel Agent-Based Framework for Analyzing Urban Agricultural Supply Chains

Abstract

In this work, we introduce a hybrid Agent-Based Simulation (ABS) andDiscrete Event Simulation (DES) model to analyze the vegetable trade dynamicsin the metropolitan city of Santiago,Chile. Our approach modelsinteractions between farmers, fairground workers or vendors, and consumers,capturing key decision-making processes in the supply chain. Farmers determinecrops based on market conditions and environmental factors, whilevendors and consumers make purchasing decisions influenced by seasonality,pricing, and availability. Given the complexity of these interactions, wepropose an optimistic approximate parallel simulation approach to enhancecomputational efficiency. Additionally, we implement a credit-based loadbalancing algorithm to address workload distribution challenges arising fromheterogeneous communication patterns among agents.