Modelling and Simulation of Road Traffic Under Rainy Conditions

Abstract

The Rainfall reduces the visibility on roads and pavement friction, which in turn increases accident risk and reduces the vehicle speed. Further, in countries where the drainage system is not proper, water collects on the streets, resulting in waterlogging, lane obstruction and the partial or complete submersion of lanes. In this paper, we study the effect of rainfall on traffic mobility using two parameters - rainfall intensity and waterlogging on the road. We conducted an image-based survey to study the drivers' perception about vehicle speed under varying rainy conditions. A total of 151 people participated in the survey. Using the traffic survey data, we designed two regression models - linear and quadratic - that represent the vehicle speed as a function of rainfall intensity and the depth of waterlogging on road. We integrated these models into the Krauss car-following model of the Simulation of Urban MObility (SUMO) to simulate the effect of rainfall on the roadway traffic. The simulation results show a very good match with the traffic survey data. We observe the Root Mean Squared Error (RMSE) of 3.8% and 1.53% for the quadratic and linear regression models, respectively. Hence, it can be concluded that the upgraded Krauss car following model can be used for traffic simulation under the rainy weather conditions.