Prediction Model of Lane Change Frequency based on Traffic Characteristics of Urban Street

Traffic Analysis and prediction

Abstract

The understanding of traffic characteristics and operation mechanisms of driving behaviors is
important for management and traffic operation strategies and planning for the future. This research focused
on investigating and predicting the effect of lane change on surrounding traffic characteristics. A Multiple
regression model of R2=0.636 was developed to estimate the frequency of lane change as a function of traffic
speed and flow rate variables. Lane change frequency (LC) was the dependent variable whereas the traffic
speed (Speed) and flow rate (Flow) were the independent variables for the model. Traffic speed influenced
the rate of lane changing more than other variables. The analysis showed a better prediction of operational
capacity depending on local and lane change frequency increasing the operational capacity by 10%. The
rate of lane changing increased as the traffic flow rate increased, and these increments were accompanied
by an increase in vehicle density with a peak lane change frequency of (1000 LC/km/hr) at a flow rate of
(1501 veh/hr) and beyond this value of flow rate the frequency decreased significantly.