Route Travel Time Reliability Model Based on Lognormal Sum Distribution

Establishing a feasible and practical route travel time reliability evaluation model is very important for accurately assessing the road network performance and providing the comprehensive traveler information.A route travel time reliability model is developed based on the probability distribution characteristics of a sum on lognormal random variables.First,the link travel time distribution model is developed resulting in that a lognormal distribution shows the best fitting performance.Second,the lognormal sum distribution-based route travel time reliability model is developed by utilizing the statistical theory and considering both scenarios of correlated versus independent links.Finally,the proposed model is validated by using the West Third-Ring road of Beijing in the case study.The results provide that the developed model can accurately evaluate the route travel time reliability,and that there is a faint correlative relationship between links,in which the best feasible correlation coefficient of links is in the range of 0.2 and 0.3.