1 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 1 / 1
Yayın Impact of traffic incident duration and road characteristics on traffic flow performance(Işık Üniversitesi, 2020-06-16) Alan, Ali Rıza; Kesten, Ali Sercan; Işık Üniversitesi, Fen Bilimleri Enstitüsü, İnşaat Mühendisliği Yüksek Lisans ProgramıNon-recurrent events that occur in any part of the highway such as accident, vehicle breakdown,fire, scattering of substances like load, fuel oil etc. on the road which disrupt the traffic flow are defined as traffic incidents. Traffic Incident Management is the planned and coordinated utilization of all human and institutional resources in order to reduce the negative effects of these incidents and to ensure the safety of the drivers, pedestrians, all other victims and response teams. Incidents generally causes congestion and yields in increased travel time which imposes costs to road users, either economic loss or reduced quality of life and mobility. This thesis presents the recent literature review on the traffic incident management theory and practice along with the traffic incident simulation models. The thesis study is comprised of two parts; a data collection and analysis and a simulation experiment phase. Firstly, Traffic incident data from Istanbul Metropolitan Area is collected and analyzed. Secondly, a microscopic traffic simulation model was developed to simulate different types of incidents and traffic demand and also various geometric characteristics of highway. The analyses are performed taking total travel time, queue length and the speed as performance measures. Lastly, to evaluate and compare the performance of the cases, statistical tests are applied. The results show that there is a linear relationship between the incident duration and the average flow speed. The increase in incident duration causes the decrease in average speed up to 11.8% for overall traffic flow and up to 29.6% for the traffic flow of the post-incident process. The longest average queue length is observed in incidents in the middle lane. The shortest average queue length is observed in incidents in the right lane. In incidents in the middle lane, queue length increases up to 34.6% compared to the incidents in the right lane for overall traffic flow. In cases where the number of vehicles per lane is the same, the increase in the number of lanes causes a decrease in average speed up to 6.4% for the traffic flow of the post-incident process. According to the results of the applied statistical tests, when overall simulation periods are analyzed, statistically different results are seen only in incident duration, vehicle input, and lane width changes. When post-incident processes are analyzed, statistically different results are seen only in incident duration and lane width changes. The results of this thesis could be used by the traffic control authorities to reduce incident duration, congestion, secondary incidents, and the associated human and economic losses.
