研究生: |
陳南宏 Nan-hung Chen |
---|---|
論文名稱: |
運用類神經網路於住宅空間格局之相似度判斷 Using neural network to judge similarity of housing spatial layout planning |
指導教授: |
施宣光
Shen-guan Shih |
口試委員: |
陳珍誠
Chen-cheng Chen 簡聖芬 Sheng-fen Chien |
學位類別: |
碩士 Master |
系所名稱: |
設計學院 - 建築系 Department of Architecture |
論文出版年: | 2006 |
畢業學年度: | 94 |
語文別: | 中文 |
論文頁數: | 27 |
中文關鍵詞: | 類神經網路 、空間格局辨識 、空間設計 |
外文關鍵詞: | Recognizing Spatial Layouts, Neural Network, Spatial Design |
相關次數: | 點閱:244 下載:4 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
許多設計師經常利用案例在室內設計的過程中輔助業主與設計師溝通。而在利用案例過程中必須進行案例擷取,即從過往的工程案例中,挑選出最具有參考價值的案例,提供給業主或設計師做為決策參考與溝通媒介。
因此本研究的目的是探討類神經網路對於擷取案例上的可行性,即利用類神經網路Artificial Neural Network辨識與篩選相似的空間格局作為設計師參考的案例。而探討的方式則是利用研究所設計的方法,以類神經網路來學習專家挑選案例的能力,再以專家的相似度判斷作為標準,評估類神經網路的辨識能力。
在研究設計的方法上,首先設計空間格局輸入的方式,再讓專家提出相似度判斷的方式與結果,類神經網路學習專家的相似度判斷和作出相似度預測,最後將專家的相似度與類神經網路的預測做比較,以比較後的準確率證明類神經網路的預測能力優劣判斷。在類神經的學習與預測的過程中,為了配合類神經網路學習的需要設計增加案例的方法,使得類神經的學習具有效率,並得以對未曾學習過的案例進行預測。
在評估類神經網路的篩選能力過程中,將利用比對與分析相似度較高的兩個案例,證明相似度高的案例在溝通設計的過程中有相當程度的參考價值。其比較的方法是比對兩個案例是否可以相互參考對方的空間格局的設計,或者直接轉換成相似案例的空間格局。
藉由上述的案例分析研究將可證明類神經網路可以辨別各空間平面的相似度,而且接近專家相似度判斷的水準。
Many interior designers use cases to aid communication with the clients. The reuse of previous cases requires storing and organizing information of prior cases, and an efficient process to retrieve adequate cases for reference in the communication. Prior cases are useful information for decision making in housing refurbishment. Cases with similar layouts are valuable information resources for non-professional housing owners to derive a realistic image over the quality, the cost and the process of the project. Recognition of housing layout patterns is an important and interesting issue in research.
This thesis describes the approach of using neural network to retrieve cases with similar layouts, which is an important step for retrieving similar cases for the references within design communication. This research proposes a method to encode the spatial layout of a house according to the orientation of functional spaces and the types of indoor-outdoor interfaces. Pairs
of encoded layout are input to a three-layer neural network for training. After the training converges to a satisfactory status, other cases are input to the network for testing. The test shows that the output of the neural network is very close to the output of a similarity function devised by a design expert.
The result confirms that neural network can be used to retrieve similar cases for design communication between interior designers and users. It is expected that with the neural network, a case- based system for supporting communication in a housing refurbishment project can be derived.
參考文獻
[1] Akin O. and Moustapha H. (2004). Strategic use of representation in architectural massing. Design study, 25, 31-50.
[2] Anthony G. O. Y. and Xun, S. (2001). Case-based reasoning (CBR) in development control. JAG. Vol. 3-Issuse 3. 238-251.
[3] Bgenstatter, U. (2000). Prediction and optimization of life-cycle costs in the early design. Building research and information, 28, 376-386.
[4] Cagdas G. (1996). A shape grammar model for designing row-houses. Design Studies, 17, 35-51.
[5] Chen A. (2004). Computational analysis on the logic of Taiwanese traditional Houses architecture. Department of Architecture and Interior Design of National Yunlin University of Science and Technology.
[6] Chen C. (2002). 『Interactive space』 A study on integrated design of information and architectural spaces. Department of Architecture of National Cheng Kung University.
[7] Chiu, M. L. (2003). Design moves in situated design with case-based reasoning. Design Studies, 24, 1-25.
[8] Chu C. T. (1997). A Study of Interior Space Planning of Apartments in Taiwan - Based on the Cases Presented by Interior Design Magazines in Taiwan. Department of Interior Design of Chung Yung Christian University.
[9] Dikmen, I., Birgonul, M. T., and Kiziltas, S. (2005). Prediction of Organizational Effectiveness in Construction Companies. Journal of Construction Engineering and Management, Vol. 131, No. 2, 252–261.
[10] Ekpar, F., Yoneda, M., and Hiroyukihase. (2003). Correcting distortions in panoramic images using construction neural networks. International Journal of Neural Systems, Vol. 13, No. 4, 239-250.
[11] Emsley, M. W., Lowe, D. J., Ryduff, A., Harding, A. and Hickson A. (2002). Data modelling and the application of a neural network approach to the prediction of total construction costs. Construction Management and Economics, 20, 465-472.
[12] Goh B. H. (2000). Evaluating the performance of combining neural networks and genetic algorithm to forecast construction demand: the case of the Singapore residential sector. Construction Management and Economics, 18, 209-217.
[13] Han, M. Cheng L. and Meng, H. (2003). Application of four-layer neural network on information extraction. Neural network, 16, 547-553.
[14] Hsiao, S. W. and Huang, H.C. (2002). A neural network based approach for product form design. Design Studies, 23, 67-84.
[15] Kim, G. H., An, S. H., Kang, K. I. (2004). Comparison of construction cost estimating models based on regression analysis, neural networks, and case-based reasoning. Building and Environment, 39, 1235-1242.
[16] Kumaraswamy, M. M. and Dissanayaka S. M. (2001). Developing a decision support system for building project procurement. Building and Environment, 36, 337-349.
[17] Ku S. H. (2003). A Research on Fashion Forecast inter Formal Manipulation in Architecture. Graduate Institute of Architecture of Tainan National University of the Arts.
[18] Lai T. T. (2002). A Study on the Characteristics of Automobile Shop and Garage-Top Five Automobile Manufacturer in Taichung City. Department of Architecture Tung-Hai University.
[19] Lao T. Q. (2004). Development of an Expert System for Interior Design. Industrial Design of National Cheng Kung University.
[20] Lee, K. J. (2000). Sufficient search space for spatial expert systems. Expert Systems with Applications, 19, 1-8.
[21] Liggett, R. S. (2000). Automated facilities layout: past, present and future. Automation in Construction, 9, 197-215.
[22] Ling, F. Y. Y. and Liu, M. (2004). Using neural network to predict performance of design-building project in Singapore. Building and Environment, 39, 1263-1274.
[23] Liu, M. and Yng, L. Y. (2003). Using fuzzy neural network approach to estimate constractor’s markup. Building and Environment, 38, 1303-1308.
[24] Nagaty, K. A. (2003). On learning to estimate the block directional image of a fingerprint using a hierarchical neural network. Neural Networks, 16, 133–144.
[25] Patricia, O. A., Eric J. R., Serafin M. D., and Cesareo E. R. (2005). Space syntax as a determinant of spatial orientation perception. International Journal of Psycholgy, 40(1), 11-18.
[26] Rivard H. and Fenves S. J. (2000). SEED-Config: A case-based reasoning system for conceptual building design. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 14, 415-430.
[27] Rumelhart D. E. and McClelland J. L. (1986). Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Vol. 1. Cambridge, MA: MIT Press.
[28] Theo J. M. V. D. V., Dick V., and Herman B. R. V. W. (1997). Comparative floorplan-analysis in programming and architectural design. Design Study, 18, 67-88.
[29] Wang T. H. (2004). A Study of Residential Interiors' Spatial Images. Department of Interior Design of Chung Yung Christian University.
[30] Watson, I. (1999). Case-based reasoning is a methodology not a technology. Knowledge-Based System, 12, 303-308.
[31] 石正義 譯,漏水問題點與解決對策(建築篇),詹氏書局,台北,2004年一月。
[32] 葉怡成,應用類神經網路,儒林圖書有限公司,台北,2001年四月。
[33] 葉怡成,類神經網路模式應用與實作,儒林圖書有限公司,台北,2004年九月。