Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ
ÇѱÛÁ¦¸ñ(Korean Title) |
ÇнÀ ¼º´É Çâ»óÀ» À§ÇÑ Â÷¿ø Ãà¼Ò ±â¹ý ±â¹Ý Àç³ ½Ã¹Ä·¹ÀÌ¼Ç °ÈÇнÀ ȯ°æ ±¸¼º ¹× È°¿ë |
¿µ¹®Á¦¸ñ(English Title) |
The Design and Practice of Disaster Response RL Environment Using Dimension Reduction Method for Training Performance Enhancement |
ÀúÀÚ(Author) |
¿©»óÈ£
À̽ÂÁØ
¿À»óÀ±
Sangho Yeo
Seungjun Lee
Sangyoon Oh
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 10 NO. 07 PP. 0263 ~ 0270 (2021. 07) |
Çѱ۳»¿ë (Korean Abstract) |
°ÈÇнÀÀº ÇнÀÀ» ÅëÇØ ÃÖÀûÀÇ ÇൿÁ¤Ã¥À» Ž»öÇÏ´Â ±â¹ýÀ¸·Î½á, Àç³ »óȲ¿¡¼ È¿°úÀûÀÎ ÀÎ¸í ±¸Á¶ ¹× Àç³ ´ëÀÀ ¹®Á¦ ÇØ°áÀ» À§ÇØ ¸¹ÀÌ È°¿ëµÇ°í ÀÖ´Ù. ±×·¯³ª, ±âÁ¸ Àç³ ´ëÀÀÀ» À§ÇÑ °ÈÇнÀ ±â¹ýÀº »ó´ëÀûÀ¸·Î ´Ü¼øÇÑ ±×¸®µå, ±×·¡ÇÁ¿Í °°Àº ȯ°æ ȤÀº ÀÚü °³¹ßÇÑ °ÈÇнÀ ȯ°æÀ» ÅëÇØ Æò°¡¸¦ ¼öÇàÇÔ¿¡ µû¶ó ±× ½Ç¿ë¼ºÀÌ ÃæºÐÈ÷ °ËÁõµÇÁö ¾Ê¾Ò´Ù. º» ³í¹®¿¡¼´Â °ÈÇнÀ ±â¹ýÀ» ½Ç¼¼°è ȯ°æ¿¡¼ »ç¿ëÇϱâ À§ÇØ ±âÁ¸ °³¹ßµÈ Àç³ ½Ã¹Ä·¹ÀÌ¼Ç È¯°æÀÇ º¹ÀâÇÑ ÇÁ·ÎÆÛƼ¸¦ È°¿ëÇÏ´Â °ÈÇнÀ ȯ°æ ±¸¼º°ú È°¿ë °á°ú¸¦ Á¦½ÃÇÏ°íÀÚ ÇÑ´Ù. º» Á¦¾È °ÈÇнÀ ȯ°æÀÇ ±¸¼ºÀ» À§ÇÏ¿© Àç³ ½Ã¹Ä·¹À̼ǰú °ÈÇнÀ ¿¡ÀÌÀüÆ® °£ °ÈÇнÀ Ä¿¹Â´ÏÄÉÀÌ¼Ç Ã¤³Î ¹× ÀÎÅÍÆäÀ̽º¸¦ ±¸ÃàÇÏ¿´À¸¸ç, ½Ã¹Ä·¹ÀÌ¼Ç È¯°æÀÌ Á¦°øÇÏ´Â °íÂ÷¿øÀÇ ÇÁ·ÎÆÛƼ Á¤º¸ÀÇ È°¿ëÀ» À§ÇØ ºñ-À̹ÌÁö ÇÇÃÄ º¤ÅÍ(non-image feature vector)¿¡ À̹ÌÁö º¯È¯¹æ½ÄÀ» Àû¿ëÇÏ¿´´Ù. ½ÇÇèÀ» ÅëÇØ º» Á¦¾È ¹æ½ÄÀÌ °Ç¹° ÈÀç ÇÇÇصµ¸¦ ±âÁØÀ¸·Î ÇÑ Æò°¡¿¡¼ ±âÁ¸ ¹æ½Ä ´ëºñ °¡Àå ³·Àº °Ç¹° ÈÀç ÇÇÇظ¦ ±â·ÏÇÑ °ÍÀ» È®ÀÎÇÏ¿´´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Reinforcement learning(RL) is the method to find an optimal policy through training. and it is one of popular methods for solving lifesaving and disaster response problems effectively. However, the conventional reinforcement learning method for disaster response utilizes either simple environment such as. grid and graph or a self-developed environment that are hard to verify the practical effectiveness. In this paper, we propose the design of a disaster response RL environment which utilizes the detailed property information of the disaster simulation in order to utilize the reinforcement learning method in the real world. For the RL environment, we design and build the reinforcement learning communication as well as the interface between the RL agent and the disaster simulation. Also, we apply the dimension reduction method for converting non-image feature vectors into image format which is effectively utilized with convolution layer to utilize the high-dimensional and detailed property of the disaster simulation. To verify the effectiveness of our proposed method, we conducted empirical evaluations and it shows that our proposed method outperformed conventional methods in the building fire damage. |
Å°¿öµå(Keyword) |
°ÈÇнÀ ȯ°æ
Àç³ ´ëÀÀ ½Ã¹Ä·¹À̼Ç
Â÷¿ø Ãà¼Ò ±â¹ý
PCA
Reinforcement Learning Environment
Disaster Response Simulation
Dimension Reduction Method
PCA
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|