Using kohonen cluster to identify time-of-day break points of intersection

Abstract

The historical traffic data is in fact capable of proving abundant of information that can aid in the development of improved current traffic control. Time-of-day (TOD) control system is the most widely used one in the world with the limited funding and system maturity, so it is important to use data mining tools that demonstrate the value of traffic data to enhance the performance of TOD systems. Kohonen cluster approach is very useful for determining TOD break points for better traffic signal control within unsupervised neural network. A case study using an intersection corridor was conducted to identify TOD break points to support the design of signal timing plan by using kohonen cluster. The results of this research indicate that the proposed method can identify TOD break points successfully without deploying multiple signal timing plans on the basis of the subjective judgment. © 2013 Springer Science+Business Media New York.