Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/77772
Title: Dataset preparation for the learning-framework of autonomous robotic system
Authors: Lee, Dong Ruen
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2019
Abstract: Machine Learning is widely used in today’s context and it has shown much interest and capability in solving problems such as playing Atari games, AlphaGo and detecting objects in image or video processing. This can be applied to applications such as a robot’s navigation system by determining the correct route to get to the destination with minimum errors and overcome obstacles that the robot might face. The scope of the project will consist of conducting tests with the Husky UGV in Gazebo simulations and to prepare datasets consisting of 2D LIDAR Point Clouds, 3D position and orientation Data. These datasets will be used in the training simulations to develop a model for the Husky’s navigation system. The project will also consist of improving the efficiency of the current system based on the time taken to train and the ability to navigate through a terrain without colliding into an object. The objective of the project will be focused on using user experience from recorded joystick actions to improve the efficiency and performance of the training. This report will focus on reporting the learning process of understanding Reinforcement Learning and the algorithms explored. It will also include the results and findings of the tests and trainings conducted on the Gazebo simulations.
URI: http://hdl.handle.net/10356/77772
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
FYP Report.pdf
  Restricted Access
Final Year Project Final Report1.42 MBAdobe PDFView/Open

Page view(s)

78
Updated on Oct 24, 2021

Download(s)

6
Updated on Oct 24, 2021

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.