Publication: Simplificación de modelos multicuerpo a través de la selección de parámetros
No Thumbnail Available
Date
2022
Editor
Director
Advisor
Coordinator
Commentator
Reviewer
Illustrator
Access rights
info:eu-repo/semantics/openAccess
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Nacional de Educación a Distancia (España), Universidad Politécnica de Madrid. Departamento de Ingeniería Mecánica
Abstract
Los métodos de selección de modelos, son usados en diferentes contextos científicos para representar un conjunto de datos en términos de un número reducido de parámetros. Los modelos multicuerpo pueden ser considerados modelos paramétricos en términos de sus parámetros dinámicos y estos modelos son susceptibles de ser reducidos. Se espera que los modelos de parámetros reducidos tengan una complejidad computacional menor que el original preservando el nivel de precisión deseado. En este trabajo, se han usado varias simulaciones para definir el conjunto de datos representativos del sistema. A continuación, un conjunto de parámetros reducido es escogido. Para tal fin, diferentes heurísticas de selección de modelos, así como del error normalizado, son propuestas. Usando estas metodologías, un robot de 6 grados de libertad ha sido analizado. Se han obtenido importantes reducciones en el número de parámetros y en su coste computacional sin prácticamente comprometer la precisión del modelo.
Model selection methods are used in different scientific contexts to represent a data set in terms of a reduced number of parameters. Multibody models can be considered parametric models in terms of their dynamic parameters, such models can be reduced. These reduced parameter models are expected to have a lower computational complexity than the original and still preserve a desired level of accuracy. In this work, several model simulations are used to define the representative data set of the system. Then a reduced set of parameters is chosen. To this end, several model selection heuristics as well as normalized error heuristics are proposed in this work. Using this methodology, a 6 degree of freedom robot has been analyzed. Significant reductions in the number of parameters and in their computational cost have been obtained without much compromise in the model accuracy.
Model selection methods are used in different scientific contexts to represent a data set in terms of a reduced number of parameters. Multibody models can be considered parametric models in terms of their dynamic parameters, such models can be reduced. These reduced parameter models are expected to have a lower computational complexity than the original and still preserve a desired level of accuracy. In this work, several model simulations are used to define the representative data set of the system. Then a reduced set of parameters is chosen. To this end, several model selection heuristics as well as normalized error heuristics are proposed in this work. Using this methodology, a 6 degree of freedom robot has been analyzed. Significant reductions in the number of parameters and in their computational cost have been obtained without much compromise in the model accuracy.
Description
UNESCO Categories
Keywords
Sistemas Multicuerpo, Robótica, Selección de Parámetros, Estimación de Parámetros
Citation
Center
E.T.S. de Ingenieros Industriales
Department
Mecánica