MODELING AND CONTROL OF VEHICULAR TRAFFIC SYSTEMS USING ARTIFICIAL INTELLIGENT NETWORK

Subject Area: Electronic and communication (real time application on computing system of cyber threat detection and mitigation)

Artificial intelligence (AI) has emerged as a transformative tool in addressing complex urban challenges, including traffic congestion in densely populated metropolitan areas. This study explores the application of AI-based models for optimizing traffic systems in Nigeria, with specific reference to the Abule Egba and Ikeja areas of Lagos State. Field data on traffic light timing and vehicle queue distances were collected and used to determine time delays within the traffic network. These measurements served as inputs to two AI models: Artificial Neural Network (ANN) and Fuzzy Inference System (FIS). The ANN model was structured with two input neurons, one output neuron, and five hidden neurons utilizing a log-sigmoid activation function. The dataset was split into 70% for training, and 15% each for testing and validation using MATLAB’s ANN toolbox. The FIS model employed four triangular membership functions for each input and 150 constant membership functions for output representation. Both AI models were integrated into a traffic management simulation in SIMULINK and evaluated using R-squared values to assess performance. Results indicated that while both AI approaches achieved prediction accuracies above 99%, the ANN model demonstrated superior performance, achieving 99.953% for Abule Egba and 100% accuracy for Ikeja. These findings underscore the potential of AI-driven models in enhancing traffic control efficiency in urban environments.