Abstract: In order to address the dual challenges posed by motor vehicle tailpipe emissions and intersection delays in urban traffic, the theory and methodology of signal control optimization at intersections in the context of low emissions are investigated.In addressing the research problem, a comprehensive analysis is conducted toward the characteristics of motor vehicle exhaust emission models（e.g, COPERT,IVE） and intersection delay models, with a focus on their optimization objectives.The study found that the COPERT and IVE models have high accuracy in simulating China’s pollution status and are suitable for emission prediction under low emission backgrounds.The study’s modeling methodology is conducted at three levels: point-control, line-control, and surface-control.It encompasses a two-layer planning model, a micro-emission model, and a follow-up model.In terms of solution algorithms, heuristics such as genetic algorithms, artificial fish school algorithms, and simulated annealing algorithms are widely used, while deep reinforcement learning algorithms and their improved algorithms show significant advantages in dynamic optimization, with the optimization efficiency improved by about 15%～30%.Attention should also be paid to the integrated application of renewable energy vehicles, driverless vehicles, and intelligent network technologies, to promote the coordinated development of "people-vehicle-road-environment".This will provide theoretical support and technical guarantees for the green and low-carbon transformation of urban transportation systems.