Adaptive Traffic Signal Control by Choosing Phase Duration

An increase in the traffic flow intensity leads to increased transportation costs, inclusive of heightened travel duration and fuel consumption. Consequently, adaptive traffic control remains a central challenge in modern urban cities. Conversely, an increase in the number of data sources and their volume, alongside advancements in machine learning and artificial intelligence techniques, provide opportunities to enhance the efficiency of solving traffic control issues. One of the most common approaches to solving traffic control problems is adaptive traffic signal control at signalized intersections. To solve this problem, various control approaches are used, including choosing the next phase, keeping or changing the current phase, and choosing the duration of the next phase. This paper focuses on an adaptive traffic signal control problem denoted as adaptively selecting the duration of the next phase from a predefined sequence of phases. We propose a modified adaptive traffic signal method aimed at maximizing the weighted traffic flow passing through the intersection during the considered time interval (phase duration). Experimental studies conducted using the SUMO microscopic traffic simulation package validated the efficacy of our proposed approach.