Traffic signal control - a neural network approach

Artificial Neural Networks (ANNs) have been proven to be an important development in a variety of problem solving areas. Increasing research activity in ANN applications has been accompanied by equally rapid growth in the commercial mainstream use of ANNs. However, there is relatively little research of practical application of ANNs taking place in the field of transportation engineering. The central idea of this thesis is to use Artificial Neural Network Software Autonet in connection with Highway Capacity Software to estimate delay. Currently existing signal control system are briefly discussed and their short coming presented. As a relative new mathematical model, Neural Network offers an attractive alternative and hold considerable potential for use in traffic signal control. It is more adaptive to the change in traffic patterns that take place at isolated intersections. ANN also provides the traffic engineer more flexibility in term of optimizing different measures of effectiveness. This thesis focuses on a better quality signal control system for traffic engineering using Artificial Neural Networks. An analysis in terms of mean, variance and standard deviation of the traffic data is also presented