Sistema inteligente e otimizado para controle de semáforos de trânsito

• Autor(a) principal: Lira, Bruno
• Data de Publicação: 2023
• Outros Autores: Rangel, Guilherme, Restaino, Paulo, Couto, Victor
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da PUC_SP
• Texto Completo: https://repositorio.pucsp.br/jspui/handle/handle/32975
• topic: CNPQ::ENGENHARIAS; Aprendizado por reforço; Inteligência artificial; Semáforo; Simulador; Trânsito; Artificial intelligence; Reinforcement learning; SUMO; Traffic; Traffic light

This study addresses the implementation of an intelligent system to control traffic lights in a city, aiming to optimize traffic flow and reduce congestion. Given the problem of urban congestion, the use of fixed-cycle traffic lights has proven to be inefficient in dealing with heavy traffic. The objective is to develop a reinforcement learning model that adjusts traffic light timings based on traffic intensity on each street.Using the SUMO simulator, an urban traffic simulation was created to collect data such as waiting time and vehicle volume. These data were used to train the reinforcement learning model. The results showed that the proposed model reduced the average vehicle waiting time by half compared to fixed traffic light timings.In summary, this study presents an innovative approach to improve urban traffic flow, using data processing techniques and reinforcement learning. The results suggest that the implementation of an intelligent traffic light control system can significantly contribute to congestion reduction and improve traffic efficiency in cities, with potential implications for the quality of life of drivers and the economy of congestion-affected urban areas