³í¹®¸í |
º¸ÇàÀÚ ±â¹Ý ½Ã¹Ä·¹ÀÌ¼Ç ÁöÇ¥¼³Á¤ ¹× °ø°£±¸Á¶·Ð°úÀÇ »ó°ü¼º ºÐ¼® / Setting Metrics for Pedestrian-Based Simulation and Correlation Analysis with Metrics for Spatial Structure Theory |
¼ö·Ï»çÇ× |
´ëÇÑ°ÇÃàÇÐȸÁöȸ¿¬ÇÕȸ ³í¹®Áý, Vol.20 n.03 (Åë±Ç85È£) (2018-06) |
ÆäÀÌÁö |
½ÃÀÛÆäÀÌÁö(33) ÃÑÆäÀÌÁö(10) |
ÁÖÁ¦¾î |
ÇàÀ§ÀÚ±â¹Ý¸ðÇü ; °¡½Ã¼º±×·¡ÇÁ ; °ø°£±¸¹®·Ð ; º¼·Ï°ø°£µµ ; Ã༱µµ ; Agent Based Model ; Visual Graph Analysis ; Space Syntax ; Convex Map ; Axial Map |
¿ä¾à1 |
°ÇÃà°ø°£ÀÇ ½Çü´Â ¹°¸®ÀûÀÌÁö¸¸ ¼º´ÉÀº ¹°¸®Àû ±â´É¿¡ ´õÇÏ¿© ±× °ø°£ÀÇ ¿ëµµ¿Í ³»ºÎ¿¡¼ »ýÈ°ÇÏ´Â »ç¿ëÀÚµéÀÇ º¹ÀâÇÑ »óÈ£°ü°è¿¡ ÀÇÇØ ÀÌ·ç¾îÁø´Ù. 70³â´ë µé¾î °ø°£ÀÇ Æ¯¼ºÀ» ºÐ¼®Çϱâ À§ÇØ ´Ù¾çÇÑ °ø°£±¸Á¶ºÐ¼®ÀÌ·ÐÀÌ Á¦½ÃµÇ°í °ËÁõµÇ¾î ¿ÔÀ¸³ª ¸ñÀû¼º(µ¿¼±, ¼ºñ½º)°ú »ç¿ëÀÚÀÇ Æ¯¼ºÀ» ¹Ý¿µÇÒ ¼ö ¾ø´Â µî ¸¹Àº ÇÑ°è¿¡ ºÎµúÇô¿Ô´Ù. º» ¿¬±¸¿¡¼´Â °ø°£À» º¹Àâ°èÀÇ °üÁ¡¿¡¼ ¹Ù¶óº¸°íÀÚ ÇÏ¿´À¸¸ç, º¹Àâ°è ºÐ¼® ÀÌ·Ð Áß ÇϳªÀÎ Agent based modelingÀ» ÀÌ¿ëÇÏ¿© °¡´É¼ºÀ» ÆľÇÇÏ°íÀÚ ÇÏ¿´´Ù. ÀÌ¿¡ space syntax µ¥ÀÌÅÍ¿¡ ´ëÀÔÇÒ ¼ö ÀÖ´Â ¹Ðµµ, ÅëÇà·®, ¿¡ÀÌÀüÆ®ÀÇ Á¸¼Ó½Ã°£ µî 3°¡Áö ÁöÇ¥¸¦ Á¦½ÃÇÏ°í, ¿©±â¿¡ 8°¡Áö ÇÁ·ÎÅäŸÀÔ ¿¹½Ã°ø°£¿¡ pedestrian based discrete event simulationÀ» ¼öÇàÇÏ¿© raw data¸¦ ÃßÃâÇÏ¿´´Ù. ¶ÇÇÑ ¿ÀÂ÷ ³» µ¥ÀÌÅÍ ¼ö·Å½Ã°£ ºÐ¼®À» ÅëÇØ ÀûÁ¤½Ã¹Ä·¹ÀÌ¼Ç ½Ã°£À» µµÃâÇÏ¿´°í, »ùÇÃÀ» ÀÌ¿ëÇÑ 10Â÷·ÊÀÇ ¹Ýº¹½Ã¹Ä·¹À̼ǰú ¿¡ÀÌÀüÆ® Á¾¼Ó½Ã°£ºÐ¼®À» ÅëÇØ work flow chart¸¦ °ËÁõÇÏ¿´´Ù. ABM½Ã¹Ä·¹À̼ÇÀ» ÅëÇØ ¾òÀº ¹Ðµµµ¥ÀÌÅ͸¦ convex mapÀÇ ºÐ¼®°á°ú¿Í, ÅëÇà·®À» axial mapºÐ¼®°á°ú¿Í ÇǾ »ó°ü°è¼ö¸¦ ÅëÇÑ °ü°è¼ºÀ» ºÐ¼®ÇÏ¿© ´ÙÀ½°ú °°Àº °á°ú¸¦ ¾òÀ» ¼ö ÀÖ¾ú´Ù.1) 10Â÷·ÊÀÇ ¹Ýº¹½Ã¹Ä·¹À̼ÇÀ» ÅëÇØ emergence dataÀÇ ÀÏ°ü¼ºÀ» È®ÀÎÇÏ¿´°í, ºü¸¥ ½Ã°£ ¾È¿¡ ¿ÀÂ÷ À̳»·Î µ¥ÀÌÅÍ°¡ ¼ö·ÅµÇ´Â °æÇâÀ» È®ÀÎÇÏ¿´´Ù. 2) ¿µ¿ª¹Ðµµ(convex map)°¡ ÅëÇà·®(axial map)¿¡ ºñÇÏ¿© ³ôÀº »ó°ü°ü°è¸¦ º¸¿´´Ù. 3) ¹Ðµµ´Â ´ëĪ¼º¿¡ ¿µÇâÀ» Å©°Ô ¹ÞÀ¸¸ç, ´ëĪ±¸Á¶°¡ ºñ´ëĪ±¸Á¶º¸´Ù »ó°ü°ü°è°¡ Å©´Ù. ÅëÇà·®Àº À§»óÀû Ư¡°ú °ü·ÃÀÌ ±íÀ¸¸ç, tree ÇüÅ°¡ ringÇüź¸´Ù »ó°ü°ü°è°¡ Å©´Ù. ƯÈ÷ ¼øȯÇü ±¸Á¶´Â ABMÀÇ ÅëÇà·®°ú °ø°£±¸Á¶À̷а£ »ó°ü°ü°è°¡ ¾ø´Ù. 4) VGA´Â °ø°£±¸Á¶À̷аú À¯»çÇÑ ¼ºÇâÀ» º¸ÀÌÁö¸¸ °ÝÂ÷´Â ÀÛ¾Ò´Ù. |
¿ä¾à2 |
The reality of the architectural space is physical, but the performance of the space is based on not only the physical function of the space but also the usage of the space and the complex interrelationships of users who live in the space. Since 1970, various theories for spatial structure analysis have been proposed and verified in order to analyze the characteristics of space, but they have faced many limitations such as their inability to reflect the purposes of space (moving path, services) and the characteristics of users. In this paper, we tried to look at the space in perspective of the complex system and tried to grasp the possibility by using the Agent-based modeling which is one of the complex system analysis theory. Hence, we propose three metrics, that can be assigned to the space syntax data, such as density, traffic volume and lifetime of the agent, and then extracted raw data by performing pedestrian based discrete event simulation in the eight prototype example spaces. In addition, the optimal simulation time was derived by analyzing the convergence time of the data within the error range, and we verified the work flow chart through 10 repetitive simulations using samples and agent dependent time analysis. By analyzing the correlation between the density data obtained from the ABM simulation and the result of convex map analysis and the correlation between traffic volume and the result of axial map analysis through Pearson correlation coefficient, the following results were obtained. 1) We confirmed the consistency of the emergence data through 10 repetitive simulations and confirmed that the data converge to within an error range in fast. 2) The convex map has higher correlation than the axial map. 3) Density is significantly influenced by symmetry, and symmetric structure is more correlated than asymmetric structure. The traffic volume is related to the topological features, and the tree topology is more correlated than the ring topology. In particular, the circular structures have no correlation with ABM traffic volume and spatial structure theory. 4) VGA has similar characteristics to space syntax, but the gap was small. Four metrics of space syntax (connectivity, control value, global integration, local integration, connectivity of visual graph analysis) were used as a comparison metric with the spatial structure theory. |