Predictive Urban Traffic Flow Model using Vehicular Big Data

Abstract

Objectives: Intelligent Transportation System (ITS) is raising the interests of the research community in the field of communication. The main objectives of this work are to implement ITS scenario and evaluate the Traffic Density Estimation for different cities for checking the density of vehicles on the different roads and perform a comparative evaluation of traffic densities in the different number of city based scenarios and to increase the accuracy of the density prediction technique for ITS. Method/Statistical Analysis: In this manner, an advanced density prediction technique has been proposed for the prediction of density on the road. The proposed technique is based on the number of vehicles on the road, in which vehicle on the map of the three cities at a particular instance of time has been estimated. After estimating the density of vehicle, prediction of that density has been done which shows that there is a maximum density on the lane and in future density on the road has been predicted. Finding: A range of values has been generated with the help of a dedicated simulation scenario using SUMO simulator. The computed values have been used for estimating the density of vehicles in Chandigarh, Delhi and Kolkata based on road side scenarios using Poisson distribution for randomly generated of vehicles and the obtained values has been evaluated to compare the densities of vehicles for these three cities. Improvement: When we are talking about traffic, then the term comes in our mind is congestion. To avoid that congestion on the roads, we use the Poisson distribution in which vehicle generates after every instance of time. With the help of this distribution, the amount of congestion on the roads can be decreased and this technique is very effective on the hurdle like jam. Keywords: Density Estimation, ITS, Linear Regression, Poisson Distribution, R Tool, SUMO Simulator, Vehicular Bigdata