Hybrid path planning for indoor robot with Laplacian Behaviour-based control via four point-explicit group

Azali Saudi and Jumat Sulaiman (2014) Hybrid path planning for indoor robot with Laplacian Behaviour-based control via four point-explicit group. International Journal of Imaging and Robotics, 12 (1). ISSN 2231-525X

[img]
Preview
Text
Hybrid path planning for indoor robot with Laplacian Behaviour.pdf

Download (47kB) | Preview

Abstract

In this paper, a robust searching algorithm to generate path planning for mobile robot is proposed. It is inspired by the behaviour-based paradigm approach to robotics architecture, in which the searching algorithm employs Laplacian Behaviour-Based Control (LBBC) during space exploration of the environment. The LBBC make use of the potential functions in the configuration space model to guide its exploration. Laplace’s equation was used to represent the potential functions in the configuration space of the robot. Consequently, the gradient of the potential functions would be used by the searching algorithm to generate path from starting to goal location. In a difficult environment, however, it suffers from the occurrence of flat region with no appreciable gradient, which result in the difficulty for the searching algorithm to generate path. The LBBC would enable the searching algorithm to generate path successfully even with the occurrence of this flat region. In this paper, the solution to Laplace’s equation is calculated via Four Point-Explicit Group (4EG) iterative method for rapid computation compared to the traditional point Gauss-Seidel iteration.

Item Type: Article
Keyword: Robot path planning, Laplacian Behaviour-Based Control (LBBC), Four Point-Explicit Group (4EG), Iterative method, Harmonic functions
Subjects: T Technology > TJ Mechanical engineering and machinery
Department: FACULTY > Faculty of Science and Natural Resources
Depositing User: MUNIRA BINTI MARASAN -
Date Deposited: 06 Aug 2018 16:28
Last Modified: 06 Aug 2018 16:28
URI: https://eprints.ums.edu.my/id/eprint/20584

Actions (login required)

View Item View Item