From Empirical Data to Operational Models: An Approach for the Development of a Decision-Making Component for an Agent-Based Mobility Simulation from Quantitative Survey Data

This contribution illustrates a holistic approach to determine and implement a decision-making component for an agent-based mobility simulation framework based on empirical data. The decision component uses a generic fuzzy inference concept developed for a flexible simulation framework for mobility scenarios, the variables and rule base of which are acquired from quantitative survey data of 194 participants. Using a minimal example case of ridesharing intention, it is shown how such a survey can be analyzed using structural equation modeling to subsequently derive a decision-making model of relevant, empirically estimated determinants for the subsequent implementation as a fuzzy inference system. The fuzzy inference system is applied in a simulation scenario to compare survey results and simulative behavior. The similar overall results of both survey and simulation regarding the used determinants indicate a general feasibility of the approach.