Application of Adaptive Neuro-Fuzzy Inference System Model on Traffic Flow of Vehicles at a Signalized Road Intersections

Abstract

In recent years, most traffic accidents and congestions usually occur at road intersections in urban areas where the vehicle speed is high. This has necessitated the need for intelligent road transport systems and high-level algorithms to unravel the problem. In this study, the South Africa Road transportation system has been used as a case study to address traffic flow solutions at signalized road intersections using traffic flow variables such as traffic density, speed of vehicles, and traffic volume as decision variables. This paper focuses on using a hybrid creative algorithm based on signalized traffic flow to address the constant repetitive traffic congestion problem. The proposed hybrid algorithm is the adaptive neuro-fuzzy inference system (ANFIS). The speed of vehicles within the investigation period, the traffic density of the road network, and the traffic volume of vehicles on the road were used as input and output variables, respectively. Triangular membership function and Gaussian membership function were used for input and output variables, and rules were developed based on available traffic flow parameters. The result of the ANFIS model showed a training and testing performance of 0.8722 and 0.9370, respectively. This training and testing results showed that the ANFIS model is an effective model for optimizing traffic flow at signalized road intersections.