Análisis de filtro de KALMAN en algoritmos de simplificación de trayectorias GPS lineales no lineales.
Fecha
2020-10
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
En la actualidad el avance de nuevas tecnolog´ıas con lo que respecta a la localizaci´on de
un objeto en movimiento, se ha podido observar un gran incremento de los datos de una
trayectoria al igual que el ruido de ellas, acumulando muchos puntos de coordenadas llegando
a redundar y a tener datos inv´alidos e innecesarios. En el proyecto se realiza el proceso de
evaluaci´on de los resultados que han sido adquiridos de las experimentaciones realizadas.
Adem´as de estudiar cual es el comportamiento que tiene el filtro de Kalman cuando es aplicado
a los algoritmos de simplificaci´on (Douglas Peucker y TD-TR) de trayectorias (GPS) lineales y
no lineales. Para esto se utilizaron 5 conjuntos de bases de datos, tantos reales como simuladas
circulares por motivos de pruebas. Para la validaci´on de los resultados se evaluaron 3 tipos
de m´etricas como son: el tiempo de ejecuci´on, margen de error y la raz´on de compresi´on.
Se concluye que al aplicar m´etodos estad´ısticos de la media y desviaci´on est´andar, donde se
obtuvieron resultados favorables con el filtro de kalman aplicados al algoritmo TD-TR, pero
adem´as se pudo observar que en tiempo de ejecuci´on el algoritmo de Douglar Peucker destaca
muchos m´as en la investigaci´on.
At present, the advance of new technologies with regard to the location of a moving object, it has been possible to observe a great increase in the data of a trajectory as well as the noise of them, accumulating many coordinate points, resulting in already have invalid and unnecessary data. In the project, the process of evaluating the results that have been acquired from the experiments carried out is carried out. In addition to studying the behavior of the Kalman filter when it is applied to the simplification algorithms (Douglas Peucker and TD-TR) of linear and non-linear trajectories (GPS). For this, 5 sets of databases were used, both real and simulated circular for testing purposes. For the validation of the results in our thesis, 3 types of metrics were evaluated, such as: execution time, margin of error and compression ratio. We concluded that statistical methods of the mean and standard deviation should be applied, where favorable results were obtained with the Kalman filter applied to the TD-TR algorithm, but we could also observe that at runtime the Douglar Peucker algorithm stands out many more in the research.
At present, the advance of new technologies with regard to the location of a moving object, it has been possible to observe a great increase in the data of a trajectory as well as the noise of them, accumulating many coordinate points, resulting in already have invalid and unnecessary data. In the project, the process of evaluating the results that have been acquired from the experiments carried out is carried out. In addition to studying the behavior of the Kalman filter when it is applied to the simplification algorithms (Douglas Peucker and TD-TR) of linear and non-linear trajectories (GPS). For this, 5 sets of databases were used, both real and simulated circular for testing purposes. For the validation of the results in our thesis, 3 types of metrics were evaluated, such as: execution time, margin of error and compression ratio. We concluded that statistical methods of the mean and standard deviation should be applied, where favorable results were obtained with the Kalman filter applied to the TD-TR algorithm, but we could also observe that at runtime the Douglar Peucker algorithm stands out many more in the research.
Descripción
PDF
Palabras clave
Simplificación, Ruido, Simulada, Filtrado, Métricas, Simplification, Noise, Simulated, Filtering, Metrics