Evaluation of Traffic Burstiness Using Gap-Based Microscopic Modelling

Abstract

Quality of traffic flow is a precondition of the business competitiveness. Smooth delivery of customers or goods increases the company competitiveness. Analysis of overcrowded roads and traffic routes that impose major problems for traffic planning teams requires the evaluation of traffic burstiness. In many everyday situations such as rush hours in the morning or evening, the traffic appears bursty that creates so-called bottlenecks. Here the microscopic modelling of stochastic processes with measurable parameters has to be implemented in order to find optimization approaches for avoiding or smoothing bottlenecks. An efficient but simple way to model bursty traffic is to use gap distribution functions describing statistically the gaps (distance) between the neighbouring cars. However, suitable gap distribution functions have to be found as the common assumption of an exponential distribution of the gaps does not allow for any bursty traffic modelling. In this work a traffic model based on gaps is introduced, and ways to estimate the parameters of the stochastic system such as the car intensity and the car burstiness are given. The model is validated by a typical traffic light situation. Our results for a given car intensity and car burstiness in a street segment with one obstacle show that a longer green phase and a shorter red phase of the traffic light is an efficient way to decrease car burstiness.