Balance of Triple Leftturn Lane Volumes at Urban Intersection and Its Influential Factors

Abstract:

A method to model and analyze balance of multiple lane volumes was proposed based on the relationship between multilane traffic distribution and driver’s lane choosing behavior. The triple leftturn lanes at urban signalized intersection were taken as case study. The collected leftturn lane volumes were organized as compositional data，and seven factors were taken into account as they could influence traffic distribution on triple leftturn lanes. A compositional analysis of variance model was formulated in the simplex to satisfy the constraint of the lane volume ratios. The dependent and independent variables are triple leftturn lane volume composition and selected influential factors respectively. The stay in the simplex approach was designed to estimate the model. Results show that the model can identify the factors with significant influence on the leftturn traffic distribution and quantify the influence degree of a factor on the ratio of each lane volume with respect to total leftturn traffic. Specific setting of the length of signal cycle，the length of vehicle turning curve inside the intersection，the setting of shadowed leftturn lane，the length of upstream segment，the location of the first leftturn sign or pavement marking at upstream，or the number of through and rightturn lanes could make the traffic distribution on triple leftturn lanes achieve relative balance.