Diseño de un algoritmo de búsqueda informada mediante el simulador robótico stage
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Fides Et Ratio
Abstract
Se propone una nueva metodología para una búsqueda informada de objetos mediante un robot móvil. La metodología está basada en una combinación de un algoritmo de búsqueda en espiral y un algoritmo de búsqueda mediante un enfoque bayesiano que usa convoluciones entre "probabilidades de observación" y "máscaras de relaciones espaciales" para estimar un mapa probabilidad donde se pueda encontrar el objeto de búsqueda. Las relaciones espaciales entre objetos son representadas mediante el uso de máscaras circulares definidos como sumas ponderadas de relaciones espaciales básicas utilizando matrices de co-ocurrencia como pesos. La metodología está validada en un ambiente tipo oficina en la que existen cuatro clases de objetos comunes ("monitor", "teclado", "cpu" y "router") y se consideró cuatro relaciones espaciales básicas ("muy cerca", "cerca", "lejos" y "muy lejos"). Se realizaron un total de 500 experimentos que comparan 5 métodos de búsqueda utilizando el simulador robótico Stage. Los resultados muestran que el uso de la metodología propuesta tiene una tasa de detección mayor que un 70%, comparado a otros métodos como ser: Método de búsqueda en espiral, Método de búsqueda informada usando partículas entre otros.
A novel methodology for robots executing informed object search is proposed. The methodology is based mainly on a heuristic spiral search algorithm and a Bayesian framework that uses convolutions between observation likelihoods and spatial relation masks for estimating the probability map of the object being search for. The spatial relationships between objects can be defined as weighted sums of basic spatial relations using co-occurrence matrices as weights. The methodology is validated in an office environment in which four object classes ("monitor," "keyboard," "system unit," and "router") and four basic spatial relations ("very near," "near," far," and "very far") are considered. A total of 500 experiments comparing five object search algorithms were realized by using the robotics simulator Stage. The results show that using the proposed methodology has a higher detection rate of 70%, compared to other methods such as: Method spiral search, method using particles informed, among others.
A novel methodology for robots executing informed object search is proposed. The methodology is based mainly on a heuristic spiral search algorithm and a Bayesian framework that uses convolutions between observation likelihoods and spatial relation masks for estimating the probability map of the object being search for. The spatial relationships between objects can be defined as weighted sums of basic spatial relations using co-occurrence matrices as weights. The methodology is validated in an office environment in which four object classes ("monitor," "keyboard," "system unit," and "router") and four basic spatial relations ("very near," "near," far," and "very far") are considered. A total of 500 experiments comparing five object search algorithms were realized by using the robotics simulator Stage. The results show that using the proposed methodology has a higher detection rate of 70%, compared to other methods such as: Method spiral search, method using particles informed, among others.
Description
Vol. 10, No. 10