A Deep Reinforcement Learning Approach for Intelligent Low-Carbon Traffic Signal Control

Improper traffic signal control will lead to long delays for vehicles and produce massive carbon emissions. The vast vehicle exhaust emissions will pollute the environment and exacerbate the earth's greenhouse effect. Traditionally, intersection signal optimization often starts from the perspective of increasing traffic efficiency but ignores reducing vehicle carbon emissions. Under the framework of a deep reinforcement learning strategy, this study proposes a novel signal control method to minimize the carbon emissions of vehicles at the intersection. To associate with carbon emissions and signal control plans, the method employs the negative value of the vehicle's carbon dioxide emissions as the reward and takes the feature vectors at different time points in the two decision action intervals as the state features. Various neural networks are adopted to extract the state features and compare their Q-value estimation effects. Numerous simulation experiments verify the effectiveness of the proposed traffic signal control model.