Hybrid calibration of agent-based travel model using traffic counts and AVI data

Artificial transportation system is critical to parallel transportation management. At its kernel is the agent-based computation which simulates individual's travel behaviors via disaggregated models and “grows” complex traffic scenarios for computational experiments. However, as a common problem, there still lacks a general calibration method for its agent travel behavioral models. Motivated by this, the paper proposes a calibration method for agent-based travel model in artificial system, which correlates macro traffic data with micro behavioral model parameters. The behavioral model is calibrated by two data sources from real urban transportation: link traffic counts and Automatic Vehicle Identification data. Our results indicate that the proposed method can help receive a reasonable model and be applied in general calibration problems.