<![CDATA[Rapid urbanization and the increasing number of vehicles in metropolitan areas have intensified traffic congestion problems, creating significant challenges for transportation efficiency, environmental sustainability, and urban mobility management in smart cities. This study aims to develop a hybrid fuzzy neural network model capable of improving traffic congestion prediction accuracy by effectively handling the uncertainty and nonlinear relationships inherent in urban traffic systems. The proposed model integrates fuzzy logic with neural network learning mechanisms to capture ambiguous traffic patterns and adaptively learn complex interactions among traffic-related variables, including traffic volume, vehicle speed, traffic density, weather conditions, time intervals, and road occupancy rates. The experimental evaluation was conducted using historical traffic datasets collected from smart city transportation environments, while model performance was assessed using Root Mean Square Error, Mean Absolute Error, Mean Absolute Percentage Error, accuracy, and the coefficient of determination. The results demonstrate that the proposed hybrid fuzzy neural network model achieved superior predictive performance and greater robustness compared with conventional machine learning approaches, particularly in scenarios characterized by dynamic traffic fluctuations and uncertain environmental conditions. Furthermore, the integration of fuzzy reasoning significantly improved the model's capability to represent gradual transitions between congestion levels and reduce sensitivity to noisy traffic observations. The study concludes that hybrid fuzzy neural network approaches provide an effective and scalable solution for intelligent traffic congestion prediction and can support data-driven traffic management strategies in future smart city ecosystems. ]]>
Copyrights © 2026
