Implementing innovative traffic simulation models with aerial traffic survey

摘要

Traffic data and human driver behaviours simulation are two of the most important parameters for the proper implementation of a traffic simulation. In our case study, we implemented an innovative method to obtain both a complex set of data of OD matrix and detailed human driver behaviours data in order to set a specific scenario simulation. Firstly, we recorded an aerial video and, subsequently, we conducted an advanced traffic analysis of aerial video data using DataFromSky service. The result is a complex complete set of traffic parameters. Data collection is based not only on classic OD matrix (volume data, vehicles classification, hourly rates, etc.), but also on innovative dynamic vehicle database parameters (speeds, lateral and tangential accelerations, travel times and distances, trajectories, etc.). All of these parameters are supplied to each vehicle detected. For this case study, our team provided as well a set of specific simulation parameters, such as gap time and follow-up time, for each entry/exit analysed. It is clear that most of this data were used to the calibration and improvement of our simulation network. In particular, we inserted specific values for each vehicle class (dimensional and dynamic values) and a detailed calibration of our behavioural models in Krauss settings. The resulting traffic simulation scenario shows the highest correlation value between real and simulated driver behaviours, probably, never obtained before.