|
|
|
Çѱ¹Áö¹Ý°øÇÐȸ / v.9, no.3, 1993³â, pp.67-76
|
AR ¸ðµ¨À» ÀÌ¿ëÇÑ »ê»ç¸é¿¡¼ÀÇ ÁöÇϼöÀ§ ¿¹Ãø
( Prediction of Groundwater Levels in Hillside Slopes Using the Autoregressive Model ) |
| ÀÌÀθð;¹Ú°æÈ£;ÀÓÃæ¸ð; °í·Á´ëÇб³ °ø°ú´ëÇÐ Åä¸ñȯ°æ°øÇаú;°í·Á´ëÇб³ ´ëÇпø Åä¸ñȯ°æ°øÇаú Á¹¾÷, ¸ñÆ÷Àü¹®´ëÇб³ Åä¸ñ°ú;;
|
|
|
 |
|
| |
| ÃÊ ·Ï |
| ¿ì¸®³ª¶ó´Â ¸¹Àº »ê¸·Áö¿ªÀ¸·ÎÀÌ·ç¾îÁ® ÀÖÀ¸¸ç ¿ì±â¿¡ ¸¹Àº»ê»çÅÂÀÇ ¹ß»ýÀ¸·Î ÀÎÇÏ¿© Àθí°ú Àç»êÀÇ ¼Õ½ÇÀ» ÀÔ°í ÀÖ´Ù. µû¶ó¼, »ê»çÅÂÀÇ ¹ß»ý¿¡ ´ëÇÑ ¿¹Ãø ½Ã½ºÅÛ°ú À§Çèµµ ºÐ¼® ¿¬±¸°¡ ÇÊ¿äÇϸç, º» ¿¬±¸ÀÇ ¸ñÀûÀº °üÃøµÈ ÁöÇϼöÀ§ÀÇ ºÐ¼®À» ÅëÇÏ¿© »ê»çÅ ¹ß»ýÀ» ¿¹ÃøÇÏ´Â °¡´É¼º¿¡ ´ëÇÑ °ÍÀÌ´Ù. À̸¦ À§ÇÏ¿© AR ¸ðµ¨À» »ç¿ëÇÏ¿© ¸ðµ¨°è¼ö¸¦ ÀÏÁ¤ÇÏ°Ô ÇÏ´Â °æ¿ì¿Í º¯È½ÃŰ´Â °æ¿ì·Î ³ª´©¾î ºÐ¼®ÇÏ¿´´Ù. AR¸ðµ¨°è¼ö¸¦ ÀÏÁ¤ÇÏ°Ô ÇÏ´Â °æ¿ì¿¡´Â AR(1), AR(2), AR(3) ¸ðµ¨À» ¼±ÅÃÇÏ¿© °¢ °¢ÀÇ ¸ðµ¨°è¼ö¸¦ ±¸ÇÏ¿´°í, AR¸ðµ¨°è¼ö¸¦ º¯È½ÃŰ´Â °æ¿ì¿¡´Â º¯ÇüµÈ AR(1)°ú ÀüÇüÀûÀÎ AR (2) ¸ðµ¨À» °úÁ¤ ¸ðµ¨·Î ÀÌ¿ëÇÏ¿© Kalman Filtering ±â¹ý¿¡ ÀÇÇÏ¿© ¸ðµ¨°è¼ö¸¦ ±¸ÇÏ¿´´Ù. ±× °á°ú, ¸ðµ¨°è¼ö¸¦ º¯È½ÃŰ´Â ½Ç½Ã°£ ¿¹Ãø ¹æ¹ýÀ̳ª AR¸ðµ¨°è¼ö°¡ ÀÏÁ¤ÇÑ °æ¿ì ¸ðµÎ »ê»ç¸é ¿¡¼ÀÇ ÁöÇϼöÀ§¸¦ Àß ¿¹ÃøÇØÁÖ¸ç, ÁöÇϼöÀ§ »Ó¸¸¾Æ´Ï¶ó ½Ã°£º° °¿ì°µµ¸¦ °í·ÁÇÔÀ¸·Î½á ´õ¿í Á¤ È®ÇÑ ¿¹ÃøÀ» ÇÒ ¼ö ÀÖÀ» °ÍÀ¸·Î »ç·áµÈ´Ù. |
|
| Korea being composed of a number of mountains has been damaged and destroyed in lives and properties by the occurrence of many landslides during the wet seasons. Therefore, it is necessary to study the forecast system and risk analysis for the occurrence of landslides : the rise of groundwater levels due to rainfall is the main cause of landslides. In this paper, the autoregressive models are used to predict the grondwater levls using cases of both time invariant and time -varing autoregressive coefficients. In the former case, AR(1), AR(2), and AR(3) models are selected and their single-valued parameters are estimated to fit them to the observed groundwater level series. In the latter case, modified AR(1) and typical AR(2) models are used as process model and a discrete Kalman Filtering technique is utilized to estimate the parameters which are themselves a function of time. The results show that the real time forecast system using the time-varying autoregressive coefficinets as well as time -invariant AR model is good to predict the groundwater level in hillside slopes and we might get better result if we use the time-hourly rainfall intensity as well as the observed groundwater level. |
| |
| Ű¿öµå |
|
| |
|
|
 |
|
Çѱ¹Áö¹Ý°øÇÐȸÁö:Áö¹Ý / v.9, no.3, 1993³â, pp.67-76
Çѱ¹Áö¹Ý°øÇÐȸ
ISSN : 1229-215X
UCI : G100:I100-KOI(KISTI1.1003/JNL.JAKO199311920445100)
¾ð¾î : Çѱ¹¾î |
|
| ³í¹® Á¦°ø : KISTI Çѱ¹°úÇбâ¼úÁ¤º¸¿¬±¸¿ø |
|
|
|
|
|
|