Detection and Control of Traffic Jams in Urban Environment

Urban traffic congestion is a recurring problem in most large worldwide cities. It usually arises from the sudden increase in the totality of vehicles on streets and avenues during peak hours. Even with continuous improvement of the urban transportation system, the number of automobiles tends to increase with the development of the economy, therefore, urban traffic congestion becomes a recurring problem. Urban congestion causes negative impacts on society, which can be highlighted, such as economic losses, reduced productivity, and increased carbon dioxide (CO2) emissions, leaving the drivers stuck in transit. The demand for a solution to avoid traffic jams is crystalline. In addition, this work proposes an algorithm for traffic congestion detection and minimization. The algorithm would find congestion areas and recommend alternative traffic routes from information collected in workable simulation time. The paper brings a solution to minimizing the congestion problems that affect a considerable part of the population in large urban centers. The obtained results, in medium and high density, the K Nearest Neighbors algorithm using Pareto II classifier (KNNP), reduces the average travel time, the stopped in traffic jams time, the time lost due to vehicle movement below the limit allowed on the road, and the average fuel consumption by towards 55%, 79%, 68%, and 28% respectively.