Deep Q-Network-Powered Optimization of Urban Public Transit for Sustainable Mobility and Efficiency

Abstract

Urban public transportation systems are essential for sustainable mobility by mitigating traffic congestion, decreasing fuel consumption, and lowering greenhouse gas emissions. However, route design, scheduling, and passenger load management deficiencies often limit their efficacy. This research introduces a framework powered by Deep Q-Network (DQN) for optimizing metropolitan public transportation systems, utilizing reinforcement learning (RL) to dynamically and efficiently address these challenges. The proposed system incorporates IoT-enabled sensors, including GPS trackers, passenger counters, and traffic flow monitors, to gather real-time data. A DQN algorithm utilizes this data to acquire optimum transit plans via ongoing interaction with the urban environment. The framework emphasizes critical performance criteria, including reducing journey times, enhancing schedule adherence, minimizing operating expenses, and improving passenger happiness. Simulation results demonstrate the framework’s ability to adapt to changing urban conditions and enhance traditional optimization techniques. The system achieves substantial improvements in resource efficiency and ecological sustainability through the strategic rerouting of buses, optimized passenger distribution, and real-time schedule adjustments. This research highlights the potential of deep RL to revolutionize urban transportation systems into more intelligent, efficient, and environmentally sustainable networks. Future research will investigate the incorporation of multi-agent systems to address scalability and multimodal transportation challenges.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.