Cycling at intersections: a multi-objective assessment for traffic, emissions and safety

Abstract

This paper evaluated in an integrated manner the traffic performance, pollutant emissions and road conflicts between bicycles and motor vehicles at a signalized intersection. Two alternative scenarios were examined: (1) bicycles increment and motor vehicles replacement within the cycle-fixed traffic signal; (2) replacing the existing traffic control by a conventional two-lane roundabout and evaluating the impacts of bicycles increment. For each scenario, bicycle demand was varied from 9 to 270 bicycles per hour (bph). Traffic flow and vehicle dynamic data were collected from a three-leg signalized intersection in Aveiro (Portugal). The microscopic traffic model (VISSIM) paired with an emission (Vehicle Specific Power – VSP) methodology and safety (Surrogate Safety Assessment Methodology – SSAM) model were used to assess intersection-specific operations. The fast Non-Dominated Sorting Genetic Algorithm (NSGA-II) was used to find the optimal bicycle demands. The results showed that two-lane roundabout outperformed the existing traffic control, namely in highest bicycle demand scenario (number of stops and travel time reduced in 78 and 14%, respectively; CO2, NOx and HC decreased 9, 7, and 12%, respectively). It was also found that the number of conflicts was significantly reduced (–49%) with this latter layout even in maximum bicycle demand scenario (270 bph). However, roundabout layout lead to more severe conflicts and potential crashes. The analysis showed that bicycle demands of 75, 95 and 110 bph delivered good environmental and safety outcomes for the intersection.

Keyword : bicycle, microscale modelling, multi-objective optimization, traffic, emissions, safety