|
|
|
Çѱ¹¼öÀÚ¿øÇÐȸ / v.40, no.10, 2007³â, pp.823-832
|
À¯ÀüÀÚ¾Ë°í¸®ÁòÀ» ÀÌ¿ëÇÑ °¿ì°µµ½Ä ¸Å°³º¯¼ö ÃßÁ¤¿¡ °üÇÑ ¿¬±¸(II): Àå.´Ü±â°£ ±¸ºÐ ¹æ¹ýÀÇ Á¦½Ã
( Parameter Estimation of Intensity-Duration-Frequency Formula Using Genetic Algorithm(II): Separation of Short and Long Durations ) |
| ½ÅÁÖ¿µ;±èżø;±è¼ö¿µ;ÇãÁØÇà; ¿¬¼¼´ëÇб³ ´ëÇпø Åä¸ñ°øÇаú;¿¬¼¼´ëÇб³ »çȸȯ°æ½Ã½ºÅÛ°øÇкΠBK21;¿¬¼¼´ëÇб³ ´ëÇпø Åä¸ñ°øÇаú;¿¬¼¼´ëÇб³ »çȸȯ°æ½Ã½ºÅÛ°øÇкΠÅä¸ñȯ°æ°øÇаú;
|
|
|
 |
|
| |
| ÃÊ ·Ï |
| º» ¿¬±¸¿¡¼´Â °¿ì°µµ½ÄÀÇ ¸Å°³º¯¼ö¸¦ ÃßÁ¤Çϱâ À§Çؼ ´Ù¸ñÀû À¯ÀüÀÚ¾Ë°í¸®ÁòÀÇ ¸ñÀûÇÔ¼ö·Î RMSE¿Í RRMSE¸¦ Àû¿ëÇÏ¿© º¸´Ù °´°üÀûÀÎ ±âÁØÀ¸·Î Àå ´Ü±â°£À» ±¸ºÐÇÏ´Â ¹æ¹ýÀ» Á¦½ÃÇÏ¿´´Ù. ¸Å°³º¯¼ö¸¦ ÃßÁ¤Çϱâ À§ÇÑ Àå ´Ü±â°£À» ±¸ºÐÇÏ´Â ¹æ¹ýÀ¸·Î´Â Á¤È®µµ¸¦ ±âÁØÀ¸·ÎÇÑ ¹æ¹ý°ú ±×·¡ÇÁ»óÀÇ Á¢Á¡À» ÀÌ¿ëÇÏ´Â ¹æ¹ýÀ» Àû¿ëÇÏ¿´À¸¸ç, ±â»óû¿¡¼ °ü¸®ÇÏ´Â 22°³ ÁöÁ¡¿¡ ´ëÇÏ¿© ±¹³»¿¡¼ ³Î¸® »ç¿ëµÇ°í ÀÖ´Â °¿ì°µµ½ÄÀ» ÀÌ¿ëÇÏ¿© ±× Àû¿ë¼ºÀ» »ìÆìº¸¾Ò´Ù. ¸Å°³º¯¼ö¸¦ ÃßÁ¤ÇÏ´Â ¹æ¹ý¿¡ À־ ÀçÇö±â°£º°·Î ¸Å°³º¯¼ö¸¦ ±¸ÇÏ´Â ¹æ¹ýÀÇ Á¤È®µµ¸¦ Æò°¡ÇÏ¿´À¸¸ç, Àå ´Ü±â°£À» ±¸ºÐÇÏ´Â ¹æ¹ý°ú ÇÔ²² ÀçÇö±â°£º°·Î ¸Å°³º¯¼ö¸¦ ±¸ÇÏ´Â ¹æ¹ýÀ» Àû¿ëÇÏ¿´´Ù. ±¹³»¿¡¼ ³Î¸® »ç¿ëµÇ°í ÀÖ´Â ¼¼ °¡Áö ÇüÅÂÀÇ °¿ì°µµ½Ä Áß¿¡¼ °¡Àå ³ôÀº Á¤È®µµ¸¦ ¾òÀ» ¼ö ÀÖ´Â °ÍÀº ÇãÁØÇà µî(1999)ÀÌ °³¹ßÇÑ °¿ì°µµ½ÄÀÎ °ÍÀ¸·Î ³ªÅ¸³µ°í, ±âÁ¸ÀÇ ¼±Çüȸ±ÍºÐ¼®À» ÅëÇØ¼ ¸Å°³º¯¼ö¸¦ ÃßÁ¤ÇÏ´Â °Íº¸´Ù ´Ù¸ñÀû À¯ÀüÀÚ¾Ë°í¸®ÁòÀ» Àû¿ëÇÑ °á°ú°¡ ´õ ³ôÀº Á¤È®µµ¸¦ º¸¿©ÁÖ´Â °ÍÀ¸·Î ³ªÅ¸³µ´Ù. |
|
| In this study, the separation of short and long durations for estimation the parameters of IDF curve is suggested by using Multi-Objective Genetic Algorithm (MOGA). Objective functions are to minimize root mean squared error (RMSE) and relative RMSE between observed and computed values. The criteria for separation are two; the first one is to estimate more precisely the parameters of IDF curve and the second is to make a single IDF curve without non-continuous duration point. For this purpose 22 rainfall recording gauges operated by Korea Meteorological Administration are selected and three IDF curves that are used generally in South Korea are tested. The result shows that the IDF curve developed by Heo et al. (1999) would be the best of three tested IDF curves, and the suggested parameter estimation method using MOGA can compute more reliable parameters compared with empirical regression analysis. |
| |
| Ű¿öµå |
| ´Ù¸ñÀû À¯ÀüÀÚ¾Ë°í¸®Áò;°¿ì°µµ½Ä;¸Å°³º¯¼öÃßÁ¤;Àå.´Ü±â°£ ±¸ºÐ;Multi-Objective Genetic Algorithm;IDF curve;Parameter estimation;Short and long duration dividing; |
| |
|
|
 |
|
Çѱ¹¼öÀÚ¿øÇÐȸ³í¹®Áý / v.40, no.10, 2007³â, pp.823-832
Çѱ¹¼öÀÚ¿øÇÐȸ
ISSN : 1226-6280
UCI : G100:I100-KOI(KISTI1.1003/JNL.JAKO200735822476574)
¾ð¾î : Çѱ¹¾î |
|
| ³í¹® Á¦°ø : KISTI Çѱ¹°úÇбâ¼úÁ¤º¸¿¬±¸¿ø |
|
|
|
|
|
|