Prototipo de un sistema de semaforización inteligente para la optimización del tráfico vehicular empleando inteligencia artificial

Abstract

This research project develops a prototype of an intelligent traffic light system to optimize vehicular traffic using an artificial intelligence model. The methodology is divided into four stages. In the first stage, vehicle flow data was collected using four cameras located at the intersection of Rodrigo Pachano Avenue and Montalvo Street in the city of Ambato. The second stage consisted of vehicle detection and counting using the YOLOv5 model and the SORT tracking algorithm, which allowed for an accurate analysis of vehicle flow. In the third stage, a data storage system with MySQL was implemented to record the number of detected vehicles. In addition, an adaptive control algorithm was developed to autonomously manage traffic light states according to the amount of traffic. Finally, in the fourth stage, a graphical interface was designed with Tkinter to supervise and control the system, and traffic was simulated with the Pygame library. A prototype using 10 mm LEDs and an ESP32 microcontroller was also integrated, which communicates with the system via the WebSocket protocol to manage the operation of the traffic lights. The results show that the system significantly improves vehicle flow, increasing traffic management capacity by 182.06%. This translates into a significant improvement in the quality of life of citizens by reducing the time needed to travel between different parts of the city.