Administración, adecuación, adaptación y pruebas de algoritmos de aprendizaje de máquinas a diferentes tipos de problemas en el sistema inteligente de control de flujo vehicular.
Fecha
2023-09
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad de Guayaquil. Facultad de Ciencias Matemáticas y Físicas. Carrera de Ingeniería en Sistemas Computacionales.
Resumen
El presente proyecto de titulación propuesto consiste en administrar, adecuar, adaptar y realizar pruebas de algoritmos que estan implementados en el Sistema Inteligente de Control de Flujo Vehicular, con una arquitectura implementada en Amazon Web Services (AWS), la cual se ejecuta en un prototipo web, donde esto se enlaza con servicios AWS tales como máquinas virtuales (EC2), almacenamiento de archivos (bucket S3), Api Gateway, funciones Lambda. Se realiza una serie de pruebas en las máquinas virtuales donde los algortimos implementados deben de soportar diferentes tipos de problemas no solo con respecto a transito vehicular, sino a otros tipos de trayectorias tales como, trayectoria de bala, trayectoria de aviones, trayectoria de una pelota de beisbol, trayectoria de movimiento de terremoto, etc. Además en este sistema se integro un nuevo modulo sobre conteo de flujo vehicular, la cual consiste que por medio de un video realiza el respectivo conteo de vehiculos, clasificandolo en buses, autos, moto, y vehiculo pesado. En conclusión se realizo pruebas por cada algortimo y con cambios en la instancia de la base de datos con la creacion de nuevas columnas y tablas se logro una mejor normalización y relación entre los datos.
Como resultados que se obtuvieron de las pruebas realizadas con los algoritmos en el sistema inteligente de control de flujo vehicular con la carga de 2 dataset y videos por algoritmos, se obtuvo los siguientes resultados: visualización de gráficos 73 cargados con éxitos y con error 13 con error y carga de dataset y videos 14 se estribaron de manera exitosa y no ocurrió ningún error.
This proposed degree project consists of managing, adapting, adapting and testing algorithms that are implemented in the Intelligent Vehicular Flow Control System, with an architecture implemented in Amazon Web Services (AWS), which runs on a web prototype, where this is linked to AWS services such as virtual machines (EC2), file storage (S3 bucket), Api Gateway, Lambda functions. A series of tests are performed in the virtual machines where the implemented algorithms must support different types of problems not only with respect to vehicular traffic, but other types of trajectories such as bullet trajectory, aircraft trajectory, trajectory of a baseball, trajectory of earthquake movement, etc. In addition, a new module on vehicle flow counting was integrated into this system, which consists of a video that performs the respective vehicle count, classifying them into buses, cars, motorcycles, and heavy vehicles. In conclusion, tests were carried out for each algorithm and with changes in the database instance with the creation of new columns and tables, a better normalization and relationship between the data was achieved. As results obtained from the tests carried out with the algorithms in the intelligent system of vehicular flow control with the loading of 2 datasets and videos by algorithms, the following results were obtained: visualization of graphics 73 loaded with success and with error 13 with error and loading of dataset and videos 14 were successfully loaded and no error occurred.
This proposed degree project consists of managing, adapting, adapting and testing algorithms that are implemented in the Intelligent Vehicular Flow Control System, with an architecture implemented in Amazon Web Services (AWS), which runs on a web prototype, where this is linked to AWS services such as virtual machines (EC2), file storage (S3 bucket), Api Gateway, Lambda functions. A series of tests are performed in the virtual machines where the implemented algorithms must support different types of problems not only with respect to vehicular traffic, but other types of trajectories such as bullet trajectory, aircraft trajectory, trajectory of a baseball, trajectory of earthquake movement, etc. In addition, a new module on vehicle flow counting was integrated into this system, which consists of a video that performs the respective vehicle count, classifying them into buses, cars, motorcycles, and heavy vehicles. In conclusion, tests were carried out for each algorithm and with changes in the database instance with the creation of new columns and tables, a better normalization and relationship between the data was achieved. As results obtained from the tests carried out with the algorithms in the intelligent system of vehicular flow control with the loading of 2 datasets and videos by algorithms, the following results were obtained: visualization of graphics 73 loaded with success and with error 13 with error and loading of dataset and videos 14 were successfully loaded and no error occurred.
Descripción
PDF
Palabras clave
AWS, PYTHON,, FLUJO VEHICULAR, MÁQUINA VIRTUAL, ALGORITMOS, VEHICULAR FLOW, VIRTUAL MACHINE, ALGORITHMS