Promoting Eco-Friendly Power and Traffic Network Operations Using Machine Learning Techniques
摘要
The growing proliferation of electric vehicles (EVs) plays a crucial role in conserving energy resources, mitigating emissions, and bolstering national energy independence. Within the expanding EV sector, commercial fleets-notably taxis and ride-sharing platforms-represent a considerable segment of the market. This study addresses this issue by proposing a standardized carbon pricing scheme that incorporates emission responsibilities to achieve low-carbon scheduling of coupled traffic-power networks. In constructing a comprehensive road network model, this research employs an origin-destination matrix derived from the gravity model and empirical road data. The study extends its analysis by simulating rapid charging requirements, formulating a framework for user-based charging decisions, and estimating waiting-associated costs through the application of queuing theory. Furthermore, it proposes a bi-level optimization scheme designed to boost the economic efficiency of both the power distribution network and its end-users. The test system shows a 5.24% reduction in emissions within the power distribution network and a 5.19% reduction in total emissions across the coupled networks.