부산대학교 도서관

The intelligent transportation systems (ITS) use a combination of sensors, cameras, and communication technologies to provide real-time data on traffic conditions. This data can be used to adjust traffic lights, but it can also be used to provide information to drivers, such as real-time traffic updates and recommended routes. Overall, control of traffic lights is an effective solution to address traffic congestion problems in urban areas. By optimizing traffic flow, cities can reduce travel times, improve safety, and reduce emissions from idling vehicles. However, the success of these systems depend on the effective planning, implementation, and ongoing maintenance to ensure that they continue to meet the changing needs of the city. In this work, the interest is to control of traffic lights in order to solve problems related to congestion; in this sense, it is proposed five mechanisms to make traffic lights intelligent and not automatic based on machine learning. The mechanisms implemented are based on four algorithms of machine learning: decision tree (DT), support vector machine (SVM), k nearest neighbor (KNN) and random forest (RF). The solutions obtained make it possible to predict the decision in order to act in traffic depending on several parameters such as the road, cars, passengers etc.