A Hybrid Dynamic Toll Pricing Framework for Congestion Management and Revenue Generation for Optimizing Urban Mobility

The need for new strategies to manage roads and increase traffic efficiency intensified as cities keep growing while traffic congestion increases. This investigation presents an entire study of hybrid dynamic toll pricing systems that manage road congestion effectively and maximizes revenue along with supporting urban environmental sustainability. Real-time adaptation of toll rates emerges from the researcher's implementation of predictive traffic models by Long Short-Term Memory (LSTM) networks together with Auto Regressive Integrated Moving Average (ARIMA) as well as Prophet to support geo-clustering techniques through DBSCAN. This system tailors toll rates to current traffic patterns alongside predicted traffic conditions. Q-learning reinforcement learning serves in this model as an automated system to establish optimal toll rate adjustments that maintain congestion reduction performance together with revenue targets. The trial simulations demonstrate that tow system produces multiple advantages such as better flow reduction combined with revenue growth as well as environmental advantages through emission decreases. Subsidy programs are suggested to support low-income passengers since the research evaluates the system's social implications. The research enables policy and transportation planning communities to create sustainability-based mobility guidelines through its provision of applicable information regarding enhanced road resource optimization and urban transportation systems.