Prediction and modeling of traffic flow of human-driven vehicles at a signalized road intersection using artificial neural network model: A South African road transportation system scenario

Abstract

In recent years, an increasing number of models and algorithms based on time series prediction and machine learning have been applied to traffic flow prediction, with good results achieved. However, few transportation researchers have been able to draw on any systematic research into the modeling of traffic flow at signalized road intersections. In this study, an artificial neural network (ANN) predictive model was developed using traffic flow variables from the South African Road transportation system as a case study. Four hundred and thirty-four (434) traffic datasets were obtained from seven road intersections connected to the N1 Allandale interchange (the busiest road in South Africa in terms of traffic volume) using sophisticated traffic data equipment such as inductive loop detectors, video cameras, and geographical positioning equipment. These traffic datasets were categorized into thirteen inputs and one output for developing the ANN model. These traffic data sets comprise of several different classes of vehicles, the speed of each category of vehicles on the road, traffic density, time, and traffic volume as input and output variables. The results obtained from this research study have shown an ANN model training and testing performance of 1.0000 and 0.99975. The evidence from this study suggests that the ANN predictive approach proposed could be used to predict and analyze traffic flow with a relatively high level of accuracy. Another significant evidence from this study suggests that the ANN model is an appropriate predictive model in modeling vehicular traffic flow at a signalized road intersection.