CONSTRUCTION OF VEHICLE DRIVING CYCLES ADAPTED TO DIFFERENT ONE-WAY TRAFFIC CONDITIONS FROM VIDEO FOOTAGE
driving cycles, fundamental diagram, microscopic traffic model.
The representative behavior of vehicles in urban roads is relevant for researchers, manufacturers, and government define technical aspects that influence the vehicle’s performance, fuel consumption, and emissions. This study aims to contribute in the comprehension of the representative behavior, considering that a track may be evaluated from a macroscopic and microscopic point of view. In addition, it is evaluated how many traffic conditions a track can present, developing a specific driving cycle for each condition. In the macroscopic evaluation, video footage from Via Mangue, an express road in Recife, measured vehicle flow and density for 276 minutes. It was obtained a fundamental diagram with vehicle capacity = 1547 cars h-1 lane-1 and critical density = 28 cars km-1 lane-1. To evaluate the track in a microscopic approach, a traffic model (Model Gipps-Santos) was calibrated from the adjust of eight input parameters, yielding an error of 2% regarding the flow and 7,5% regarding the vehicular energy demand. From the fundamental diagram and microscopic model, it was verified that the track can be characterized in three distinct conditions (free flow, intermediate flow and congested flow) within the average error of 5,0%. The three driving cycles were developed through the Markov Chain-Monte Carlo method, with average density 15.6/54.9/92.3 cars km-1 lane-1, average deviation of 3.5%/3.9%/3.5% with regard to characteristic parameters and 0.29/0.36/0.35 MJ km-1 of vehicular energy demand. The three developed driving cycles were compared to other local driving cycles around the world in terms of its kinematic parameters and energy demand. It was also found that employing just one driving cycle to represent every track condition could provide up to 35% of error.