Abstract:
This paper suggests a model to predict Average Daily Traffic (ADT) data of road segments in Sri Lanka where timely updated ADT data are not available. This methodology can be used to predict and estimate ADT data of both major and minor roads. Many previous studies have been conducted to estimate ADT only on specific road segments instead of considering the whole road network in the country. Initially, a road segment with a considerable length is selected for the study, and a model is developed to find the relationship between available ADT data of the particular road segment and associated factors which influence the ADT. These factors include social factors, economic factors, roadway and land use characteristics, availability of public transport, infrastructure development, etc. The most significant variables that affect ADT are identified, and corresponding weights are assigned for each variable through the developed regression model. Predicted accuracy of the model is validated through manual traffic counts. This proposed methodology can be applied for both major and selected minor roads and then expanded to the entire road network. This prediction model can be used to estimate ADT of road segments where updated ADT is not available, and data can effectively be used for transportation planning, capacity analysis, and infrastructure design, etc.
