This study aims to develop an innovative method to generate a traffic control system on the basis of reinforcement learning. Afterward, its performance is computed concerning WT (waiting time) and SD (Standard Deviation). The results exhibited that the developed method is more effective and superior to traditional methods and leads to mitigate WT and SD. Lower WT indicates that the developed method is effective to manage traffic flow, due to which the congestion is alleviated and enhanced traveling experience is obtained. Additionally, the least SD results in illustrating a more expectable traffic pattern, which makes the travel times more secure and reliable. These findings are useful to enhance metropolitan flexibility and lessen the ecological impact of traffic jamming. The Reinforcement Learning (RL) assists the urban areas in deploying more sustainable transportation schemes, that are helpful for commuters and the setting.
Improved Reinforcement Learning Model for Traffic Management
2024-08-23
729371 byte
Conference paper
Electronic Resource
English