An Agent-Based Modeling Approach for Transportation Network Systems

This paper presents the ground work necessary to study the evolving air transportation industry. A survey of related models based on a system-of-systems perspective results in a methodology formulated to address overlooked areas such as market coordination and competition. An agent-based modeling approach underpins the integration of the demand-side (consumers) and supply-side (service providers) entities to effectively capture the dynamic interactions between these stakeholders within the National Transportation System. Machine Learning techniques are adopted to model the competitive sentience of air service providers in making decisions that affect consumers’ travel demand over time and vice versa. An interim model is developed employing this methodology on a four-node, inter-city transportation network, and has shown promising results toward further development efforts.