-
Notifications
You must be signed in to change notification settings - Fork 0
/
2021-arw-open-loop-robot-control-using-dqn.html
173 lines (150 loc) · 8.28 KB
/
2021-arw-open-loop-robot-control-using-dqn.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
<!DOCTYPE html>
<html>
<head>
<title>Simon Schwaiger - Github IO</title>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=0.9">
<link rel="stylesheet" href="https://www.w3schools.com/w3css/4/w3.css">
<link rel="stylesheet" href="./src/styles.css">
<link rel="shortcut icon" type="image/x-icon" href="/favicon.ico"> <!--http://tools.dynamicdrive.com/favicon/-->
</head>
<body class="w3-light-grey">
<!-- w3-content defines a container for fixed size centered content,
and is wrapped around the whole page content, except for the footer in this example -->
<div class="w3-content">
<!-- Grid -->
<div class="w3-row">
<!-- Introduction menu -->
<div class="w3-col l4">
<!-- About Card -->
<div class="w3-card w3-margin w3-margin-top">
<a href="./index.html">
<img src="./img/portrait.jpg" style="width:100%; border-top-left-radius: 20px; border-top-right-radius: 20px;">
</a>
<div class="w3-container w3-white" style="border-top-left-radius: 0px; border-top-right-radius: 0px;">
<a href="./index.html">
<h4><b>Simon Schwaiger, MSc</b></h4>
</a>
<p>I am a Lecturer/Researcher at University of Applied Sciences Technikum Wien and Doctoral Student at Graz University of Technology, working on machine learning and modern control approaches in robotics as well as personal projects.</p>
<ul style="list-style-type:none; padding:0px; margin:0px;">
<hr>
<li>
<a href="./cv.html">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/briefcase.svg"><span><b>Curriculum Vitae</b></span></b>
</button>
</a>
</li>
<li>
<a href="./publications.html">
<button class="button-highlight w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/document-signed.svg"><span><b>Publications</b></span></b>
</button>
</a>
</li>
<li>
<a href="./portfolio.html">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/square-terminal.svg"><span><b>Portfolio</b></span></b>
</button>
</a>
</li>
<li>
<a href="./teaching.html">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/graduation-cap.svg"><span><b>Teaching Experience</b></span></b>
</button>
</a>
</li>
<hr>
<li>
<a href="https://github.com/SimonSchwaiger" target="_blank">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/github.svg"><span><b>GitHub</b></span></b>
</button>
</a>
</li>
<li>
<a href="https://www.researchgate.net/profile/Simon-Schwaiger" target="_blank">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-50%,-2.5px);" src="./img/icons/book-bookmark.svg"><span><b>Researchgate</b></span></b>
</button>
</a>
</li>
<li>
<a href="https://www.linkedin.com/in/simon-schwaiger-90354519a/" target="_blank">
<button class="w3-button w3-padding-16 w3-white w3-block w3-left-align">
<img style="height:20px; transform:translate(-40%,-2.5px);" src="./img/icons/linkedin.svg"><span><b>LinkedIn</b></span></b>
</button>
</a>
</li>
</ul>
<p></p>
</div>
</div>
<!-- END Introduction Menu -->
</div>
<!-- Information Cards -->
<div class="w3-col l8 s12">
<!-- Content Class -> Holds one Publication -->
<div class="w3-card-4 w3-margin w3-white" style="padding: 15pt;">
<!-- START GENERATED HTML HERE --><h1 align="center">
Open Loop Robot Control using Deep Q-Learning
</h1>
<h3 align="center">
Simon Schwaiger<sup>1</sup>, Ali Aburaia<sup>1</sup>, Mohamed Aburaia<sup>1</sup> and Wilfried Wöber<sup>1</sup>
</h3>
<i align="center">
<p><sup>1</sup> University of Applied Sciences Technikum Wien, Faculty of Industrial Engineering, 1200 Vienna, Austria</p>
<p><a href="mailto:novotny@technikum-wien.at">schwaige@technikum-wien.at</a></p>
</i>
<table align="center" style="border-collapse: collapse; max-width: 100pt;">
<tr>
<td align="middle" style="border: none;">
<a href="https://www.researchgate.net/publication/363480276_Open_Loop_Robot_Control_using_Deep_Q-Learning" style="color: white; font-size: 14pt;">
<div style="background-color: #363636; border-radius: 50px; padding: 10px 20px; color: white; width: 80pt;">
<img src="img/document_icon.png" height="14" style="transform:translate(-10%,-1px);"> Paper
</div>
</a>
</td>
</tr>
</table>
<h2 align="center"> Abstract</h2>
<i align="center">
<p>Motion planning is essential for any robot application. Since motion planning algorithms require explicit knowledge about the configuration, constraints and physical parameters of the robot, motion planning algorithms are often specific to a certain robot configuration or robotics task. While reinforcement learning can be used to have a robot learn to achieve a task, these implementations are typically done as end-to-end solutions, where a model is trained on a real robot, taking raw sensor data as the input and outputting joint torque. For this type of system, however, training typically has to be performed on the real robot. We implement and evaluate a system that trains a neuronal network to determine robot trajectory in simulation instead of the real robot. We create a modular system for robot arm control, where joints are controlled by an open loop controller, while kinematics and trajectory planning are done using a reinforcement learning model. The model is trained in simulation using multi-goal reinforcement learning, allowing one neuronal network to plan a path to multiple arbitrary goal points. We train models using training optimisations such as reward shaping and hindsight experience replay. The results show that motion planning can be performed by the presented approach, allowing an autonomously trained neuronal network to determine robot movement and adapt to changing goals and a changing environment.</p>
</i>
<hr />
<h2>Citation</h2>
<p>If you use this work in your research, please cite our paper:</p>
<pre><code class="language-bibtex">@inproceedings{Schwaiger2021Open,
author = {Schwaiger, Simon and Aburaia, Ali and Aburaia, Mohamed and Wöber, Wilfried},
title = {Open Loop Robot Control using Deep {Q}-Learning},
editor = {Kubinger, Wilfried and Brandstötter, Mathias and Schöffmann, Christian and Vincze, Markus},
booktitle = {Proceedings of the Austrian Robotics Workshop 2021},
month = {10.-11. June},
address = {Vienna, Austria},
pages = {12-17},
year = {2021},
isbn = {978-3-200-07799-7}
}
</code></pre>
<!-- END GENERATED HTML HERE -->
</div>
<!-- END Information Cards -->
</div>
<!-- END GRID -->
</div>
<!-- END w3-content -->
</div>
<!-- Footer -->
<footer>
<p style="margin-top:3cm;"><b>2024 Simon Schwaiger.</b>
Powered by <u><a href="https://www.w3schools.com/w3css/default.asp" target="_blank">w3.css</a></u>.
Icons by <u><a href="https://www.flaticon.com/icon-fonts-most-downloaded" target="_blank">flaticon.com</a></u>.
Styling modified from <u><a href="https://github.com/KrauseFx/markdown-to-html-github-style/tree/master" target="_blank">markdown-to-html-github-style.</a></u>
Publication pages inspired by the <u><a href="https://nerfies.github.io/" target="_blank">nerfies github.io page</a></u>.</p>
<p>This website is licensed under a <u><a href="https://creativecommons.org/licenses/by-sa/4.0/" target="_blank">Creative Commons Attribution-ShareAlike 4.0 International License.</a></u>
Website source code can be borrowed, but link back to the template in your footer.</p>
</footer>
</body>
</html>