A Hybrid Optimization and Machine Learning Framework for Urban Traffic Management Using Cyber-Physical Digital Twin Architecture

As cities continue to expand rapidly, traffic congestion has become a pressing issue, necessitating advanced traffic management systems. This research proposes a CyberPhysical Digital Twin (CPDT) architecture for optimizing urban traffic and simulating smart city transportation systems in realtime. The CPDT framework integrates data from various sources to provide a dynamic and real-time overview of city traffic. In this enhanced approach, a hybrid methodology combining Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) algorithms is utilized to optimize both route planning and traffic signal timings. Additionally, advanced machine learning model Transformer networks are employed to forecast traffic patterns and incidents. These improvements lead to a 20% reduction in average traffic delays and a 15% increase in prediction accuracy, thereby enhancing traffic management outcomes and overall prediction reliability.