Rule-Based Mode Choice Model: INSIM Expert System

Abstract

This paper presents an innovative rule-based intelligent network simulation model (INSIM) expert system (IES) which simulates real-time mode choice decision-making process of commuters in the presence of multimodal traveler information. The IES captures interactions among available modes and decides on the commuter’s mode based on a commuter’s socioeconomic traits and prevailing travel condition. The commuter’s mode choice behavior is modeled and represented by cognitive rules in the rule-base of the IES. Two important characteristics of the IES, the reliability and the adaptive learning, are highlighted. Three different models, i.e., (1) pure rule-based model (PRB), (2) discrete choice model (DCM), and (3) probabilistic model (COM) are introduced to formulate the mode choice decisions. Simulation results show that the highest level of accuracy can be achieved by applying the PRB model to generate mode choice decisions.