Agrupamiento de datos en fluidos no estacionarios
Archivos
Fecha
2014-06-26
Autores
Gorrín Ortega, Yuney
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
En el área de la Visualización Científica, el análisis y visualización de fluidos ocupa una rama muy importante y atractiva. Los campos vectoriales, objeto matemático con el cual se representan los fluidos, son de uso común para representar otros fenómenos en la ciencia, principalmente de origen natural.
Se han desarrollado varias técnicas para analizar estos tipos de datos basadas en integración numérica, en texturas, en la extracción de rasgos topológicos y en visualización directa de volúmenes. Esta investigación pretende visualizar la evolución de las zonas de comportamiento similar en fluidos no estacionarios. Por tanto se opta por usar las streamlines, técnica basada en integración que por su propia naturaleza representa de forma intrínseca la geometría del fluido. Luego se aplica un algoritmo de agrupamiento que determina la cantidad óptima de grupos en cada instante de tiempo. Partiendo de estos agrupamientos se construye un grafo que representa la evolución de los grupos de un instante de tiempo a otro. Una vez que se tiene el grafo se puede analizar el fluido determinando la evolución de los grupos y visualizando las streamlines que pertenecen a cada uno de ellos.
On the field of Scientific Visualization, analysis and visualization of fluids occupy a very important and attractive area. Vector fields wich are mathematical objects used to represent fluids, are commonly used to represent other phenomena in science, mainly of natural origin. Several techniques have been developed to analyze these data types based on numerical integration, in texture, in the extraction of topological features and direct volume visualization. This research aims to visualize the evolution of the areas of similar behavior in non-stationary fluids. Therefore we choose to use streamlines, an integration based technique that by its nature inherently represents the geometry of the fluid. A clustering algorithm that determines the optimal number of groups at each time point is then applied. From these clusters we construct a graph that represents the evolution of the groups from a moment of time to another. Once we have the graph the fluid can be analyzed by determining the evolution of groups and visualizing streamlines belonging to each one of them.
On the field of Scientific Visualization, analysis and visualization of fluids occupy a very important and attractive area. Vector fields wich are mathematical objects used to represent fluids, are commonly used to represent other phenomena in science, mainly of natural origin. Several techniques have been developed to analyze these data types based on numerical integration, in texture, in the extraction of topological features and direct volume visualization. This research aims to visualize the evolution of the areas of similar behavior in non-stationary fluids. Therefore we choose to use streamlines, an integration based technique that by its nature inherently represents the geometry of the fluid. A clustering algorithm that determines the optimal number of groups at each time point is then applied. From these clusters we construct a graph that represents the evolution of the groups from a moment of time to another. Once we have the graph the fluid can be analyzed by determining the evolution of groups and visualizing streamlines belonging to each one of them.
Descripción
Palabras clave
Agrupamiento de Datos, Fluidos no Estacionarios, Estadística Computacional, Computación Gráfica