Model Study for Intelligent Transportation System with Big Data

The increasing development of big data technology has brought great opportunities and challenges to innovation of complex system such as Intelligent Transportation System (ITS), especially in the method of big data driven modelling. This paper focused on analyzing the feasibility of model study based on noisy trajectory data collected by cell phone for ITS. A method of real-time modelling based on trajectory data is proposed and specific experiment is designed for analysis. An improved GHR car-following model is accepted in this paper for parameters calibration by Least-square method. Sensitivity analysis and cross-calculation are performed to validate if the modelling process is reliable and robust enough. Research results show practicable of noisy data in data-driven modeling. The results of this paper show feasibility of using trajectory data from cell phone for dynamic modeling for Intelligent Transportation System.