Multi-Method Approach of System Dynamics And Agent-Based Modeling: A Practical Case Study of the Norwegian Dairy Supply Chain

Abstract

A multi-method, hybrid model was developed combining system dynamics and agent-based modeling for the Norwegian dairy product supply chain. The model and relevant data was adapted from an ongoing work in developing a purely agent-based model and translated into the software tool AnyLogic. The model consists largely of three agent populations: farm, processor, and municipalities. Unique stock-and-flow structures were built into each agent population, with particular attention to farms and processors. Results of the hybrid model with purely agent-based as well as system dynamic were compared. Some variables from the hybrid model were found to be comparable to the agent-based model, while others were similar to that from the system dynamics model. Moreover, there were some emergent feedback behaviors not observed in either of the single-method models.