Publication detail
MOBILE ROBOT MOTION PLANNER VIA NEURAL NETWORK
KREJSA, J. VĚCHET, S.
Czech title
MOBILE ROBOT MOTION PLANNER VIA NEURAL NETWORK
English title
MOBILE ROBOT MOTION PLANNER VIA NEURAL NETWORK
Type
conference paper
Language
en
Original abstract
Motion planning is essential for mobile robot successful navigation. There are many algorithms for motion planning under various constraints. However, in some cases the human can still do a better job, therefore it would be advantageous to create a planner based on data gathered from the robot simulation when humans do the planning. The paper presents the method of using the neural network to transfer the previously gained knowledge into the machine learning based planner. In particular the neural network task is to mimic the planner based on finite state machine. The tests proved that neural network can successfully learn to navigate in constrained environment.
Czech abstract
Motion planning is essential for mobile robot successful navigation. There are many algorithms for motion planning under various constraints. However, in some cases the human can still do a better job, therefore it would be advantageous to create a planner based on data gathered from the robot simulation when humans do the planning. The paper presents the method of using the neural network to transfer the previously gained knowledge into the machine learning based planner. In particular the neural network task is to mimic the planner based on finite state machine. The tests proved that neural network can successfully learn to navigate in constrained environment.
English abstract
Motion planning is essential for mobile robot successful navigation. There are many algorithms for motion planning under various constraints. However, in some cases the human can still do a better job, therefore it would be advantageous to create a planner based on data gathered from the robot simulation when humans do the planning. The paper presents the method of using the neural network to transfer the previously gained knowledge into the machine learning based planner. In particular the neural network task is to mimic the planner based on finite state machine. The tests proved that neural network can successfully learn to navigate in constrained environment.
Keywords in Czech
Mobile robot; motion planning; neural networks
Keywords in English
Mobile robot; motion planning; neural networks
RIV year
2011
Released
09.05.2011
Publisher
Institute of Thermomechanics, Academy of Sciences of the Czech Republic
Location
Prague
ISBN
978-80-87012-33-8
Book
Engineering Mechanics 2011, 17th International conference may 9-12, 2011, Svratka, Czech republic
Edition number
1
Pages from–to
227–230
Pages count
4
BIBTEX
@inproceedings{BUT98286,
author="Jiří {Krejsa} and Stanislav {Věchet},
title="MOBILE ROBOT MOTION PLANNER VIA NEURAL NETWORK",
booktitle="Engineering Mechanics 2011, 17th International conference may 9-12, 2011, Svratka, Czech republic",
year="2011",
month="May",
pages="227--230",
publisher="Institute of Thermomechanics, Academy of Sciences of the Czech Republic",
address="Prague",
isbn="978-80-87012-33-8"
}