Fuzzy Logic Base Large-Scale Urban Traffic Signal Optimization Using Circular Complex Pythagorean Fuzzy Sets and Intelligent Hybrid Machine Learning

Abstract

Large-scale urban traffic networks operate under highly dynamic, nonlinear, and uncertain conditions, where fluctuating traffic demand, incomplete sensor data, and unpredictable driver behavior make real-time signal optimization a complex decision-making challenge. To address these issues, this study introduces the concept of Circular Complex Pythagorean Fuzzy Sets (CrC-PFS) for modeling uncertainty in urban traffic signal control problems. The proposed CrC-PFS framework extends classical Pythagorean fuzzy sets, Complex Pythagorean fuzzy sets, and circular Pythagorean fuzzy environments by providing greater flexibility in representing membership, non-membership, and hesitation degrees under periodic and time-varying traffic patterns. To enhance computational capability, we establish refined algebraic operational laws for CrC-PFS, including direct sum, direct product, and scalar multiplication operators based on generalized t-norm and t-conorm structures. In addition, Circular Complex Pythagorean fuzzy weighted averaging and ordered weighted aggregation operators are developed to integrate multiple traffic performance indicators such as queue length, delay time, saturation flow, and emission levels within a multi-criteria decision-making framework. Furthermore, an intelligent hybrid machine learning mechanism is incorporated to dynamically learn traffic flow patterns and adapt signal timing strategies in large-scale urban networks. By integrating fuzzy uncertainty modeling with predictive learning algorithms, a robust optimization framework is constructed for adaptive traffic signal coordination. The experimental findings demonstrate that the proposed model significantly improves traffic efficiency, reduces congestion levels, and enhances overall network resilience, thereby supporting sustainable and uncertainty-aware urban transportation management.