A Simulation based Evaluation of Traffic State Estimation Accuracy by Using Floating Car Data in Complex Road Networks

Abstract:

This study developed a simulation based approach to comprehensively evaluate the joint effects of penetration rate and uploading frequency of floating car data on traffic state estimation accuracy for the large scale complex road networks. A case study based on the simulation model of Lujiazui Region in Shanghai was conducted to demonstrate the applicability of the developed approach and explore the optimal penetration rate and uploading frequency in terms of travel speed estimation accuracy and network coverage. Totally, 30 combinations of the two parameters were tested in the calibrated simulation model. Results show that the accuracy of travel speed estimation and network coverage rise as the penetration rate and uploading frequency increase. The optimal penetration rate and uploading frequency are found to be 8% and 1/10 s-1 respectively.

Key words:large scale complex road networks; floating car data; traffic state estimation; traffic simulation