Light Avoiding SBOT - Revision history

Robot at 17:43, 17 June 2012

2012-06-17T17:43:15Z

2012-06-16T11:00:33Z

2012-06-13T00:17:34Z

‎Video

2012-06-13T00:08:01Z

‎Results (this test is on youtube)

2012-06-13T00:07:47Z

‎Results (this test is on youtube)

2012-06-13T00:06:34Z

‎Results (this test is on youtube)

2012-06-13T00:06:08Z

‎Results (this test is on youtube)

2012-06-13T00:05:47Z

‎Results (this test is on youtube)

2012-06-13T00:04:37Z

‎Results (this test is on youtube)

2012-06-13T00:03:49Z

‎Results - first sensor state, second sensor state | values of rules

← Older revision		Revision as of 17:43, 17 June 2012
Line 85:		Line 85:
	==Source Code==		==Source Code==

−	Source code: [[~~{{ns:media}}~~:Light_avoiding_source.rar]]	+	Source code: [[Media:Light_avoiding_source.rar\|Light_avoiding_source.rar]]

← Older revision		Revision as of 11:00, 16 June 2012
Line 1:		Line 1:
		+	Authors: Martin Ďurčo, Dominik Kapišinský
		+
	In our project we use SBOT robot. Goal for this project is to make this robot learn the best possibility, how he could run away from light.[[Image:light_sbot_1.jpg\|thumb\|250px\|Light avoiding SBot]]		In our project we use SBOT robot. Goal for this project is to make this robot learn the best possibility, how he could run away from light.[[Image:light_sbot_1.jpg\|thumb\|250px\|Light avoiding SBot]]

← Older revision		Revision as of 00:17, 13 June 2012
Line 80:		Line 80:

	http://www.youtube.com/watch?v=KFvERq0Zsfc		http://www.youtube.com/watch?v=KFvERq0Zsfc
		+
		+	==Source Code==
		+
		+	Source code: [[{{ns:media}}:Light_avoiding_source.rar]]

@@ Line 39: / Line 39: @@
 ==Results (this test is on youtube)==
-First sensor state, Second sensor state | FORWARD value, moveLEFT value, moveRIGHT value
+First sensor state, Second sensor state | FORWARD value, moreLEFT value, moreRIGHT value
 After 1. minute - Reinforcement Learning was ON

@@ Line 39: / Line 39: @@
 ==Results (this test is on youtube)==
-First sensor state, Second sensor state | Values of rules for that sensor combination
+First sensor state, Second sensor state | FORWARD value, moveLEFT value, moveRIGHT value
 After 1. minute - Reinforcement Learning was ON

@@ Line 41: / Line 41: @@
 First sensor state, Second sensor state | Values of rules for that sensor combination
-After 1. minute - Reinforcement Learning is ON
+After 1. minute - Reinforcement Learning was ON
 ,0 | 442, 193, 268<br>
@@ Line 53: / Line 53: @@
 ,2 | 171, 197, 299<br><br>
-After 2 minutes - Reinforcement Learning is OFF
+After 2 minutes - Reinforcement Learning was OFF
 ,0 | 442, 193, 268<br>
@@ Line 65: / Line 65: @@
 ,2 | 171, 197, 299<br><br>
-After 3 minutes - Reinforcement Learning is again ON
+After 3 minutes - Reinforcement Learning was again ON
 ,0 | 505, 271, 316<br>

@@ Line 38: / Line 38: @@
 The first two of these problems could be considered planning problems (since some form of the model is available), while the last one could be considered as a genuine learning problem. However, under a reinforcement learning methodology both planning problems would be converted to machine learning problems.
-==Results - first sensor state, second sensor state | values of rules==
+==Results (this test is on youtube)==
 ,0 | 442, 193, 268<br>
@@ Line 49: / Line 52: @@
 ,1 | 211, 200, 200<br>
 ,2 | 171, 197, 299<br><br>
 ,0 | 442, 193, 268<br>
@@ Line 59: / Line 64: @@
 ,1 | 211, 200, 200<br>
 ,2 | 171, 197, 299<br><br>
 ,0 | 505, 271, 316<br>