醫學中心除了提供完整的醫療服務，也提供緊急救護、醫學研究與教學的服務，完整的醫療硬體支援著提供的醫療品質。隨著國人對醫療需求的日益增加，醫院建築也逐漸朝大型化與複雜化發展，其維護與修繕的迫切度也遠高於一般建築，而醫院的維持良好運作高度仰賴著硬體設備的維護管理。
然而，在每年國家補助預算遞減的情況下，即便如國家醫學中心都必須自負盈虧。因此，如何在有限的預算之下，尋找出關鍵維護項目，進而控管大宗維護金額花費，是本研究所探討的課題。因此，本研究之目的在於探討醫院建築的關鍵維護項目、建立維護成本預測模型以及維護方案排序決策模型，提供管理階層以更有效率的方式管理醫院建築。
在醫院建築關鍵維護項目分析方面，本研究將維護成本依其類型分為定期維護、損壞修繕與需求變更等三大類，分別探討台大醫院東西址院區在此三大類型中的關鍵維護項目與細項，並比較不同生命週期階段的維護行為差異。本研究亦整理國外文獻中之每單位醫院建築維護成本與台大醫院進行對照討論。分析結果顯示：
1. 在定期維護類別中，東址院區與西址院區的關鍵維護項目均為重電設備（東址22.1%、西址26.1%）、HVAC（東址26.2%、西址25.9%）與昇降運輸系統（東址13.9%、西址15.9%），其項目與比例均相近。由此可知，定期維護項目與醫院特性較為相關，反而與生命週期階段的關聯性較低。
2. 在損壞修繕類別中，東址院區與西址院區的關鍵細項均為內部裝修、弱電設備與HVAC，然比例上亦有所差異。東址與西址的損壞修繕關鍵細項分別為：（1）東址：內部裝修（43.4%）、HVAC（24.7%）、弱電設備（9.3%）；（2）西址：內部裝修（67.3%）、弱電設備（10.6%）、HVAC（5.1%）。由於西址為老舊建築，建築本體的體質不似青壯期的醫院，有將近七成的損壞發生在內裝、外牆、屋頂、門窗等修繕工程，且其總損壞修繕費用每年每平方米之平均花費成本為東址的2.8倍(西址為平均每年22.8美元/m2，東址為平均每年8美元/m2)。
3. 在需求變更類別中，東址院區與西址院區的關鍵細項與比例上均有所差異，其分別為：（1）東址，弱電設備（52.7%）、鍋爐蒸氣設備（31.3%）、內部裝修（11.1%）、HVAC（4.1%）；（2）西址，內部裝修（29.9%）、弱電設備（27.6%）、HVAC（17.9%）、昇降運輸系統（16.0%）。東址院區相對屬於較年輕、符合使用期待的建築，因此在需求變更上，更改使用格局或重新分配空間等內部裝修上相對較少，反而在於新增弱電設備、鍋爐蒸汽之需求較多；反觀西址屬老年化建築，因此必須大幅修改內部空間狀態以符合使用需求，而衍生出弱電、空調系統等修繕成本。
在建築維護成本預測模型方面，本研究以灰色系統理論為架構，整合指數級數與自組織映射網路等殘差修正技術，分別調整規則性殘差與不規則性殘差，以建立有效的醫院建築維護預算預測模型，預估未來一段時間內的預算需求，以提供醫院工務室於預算編列之參考。本研究以台大醫院東址院區1998年到2006年共36季的每季損壞修繕成本資料進行測試，所預測之成本與實際值平均誤差為5.12%。雖然本研究所提出的預測模型在序列轉折處仍有些許過衝現象，但最大誤差值仍在9.61%之下，顯示預測模型能快速地修正預測方向，且精確地掌握序列發展的動態，證實本研究所提出模型的實用性與有效性。
最後，在建築維護方案排序決策模型方面，本研究整合灰數操作定義、權重設定技術與關聯分析技術等，以提升決策模型處理資訊的能力與可靠度。本研究所採用的權重設定技術係基於離差最大化的概念，對於能辨識方案差異的屬性給予較高權重。此外，在關聯分析方面，本研究採用灰數灰關聯分析技術，不僅可將維護方案依理想方案為基礎進行排序，並可進一步處理不確定資訊。透過案例的說明，此模型可簡化專案評選的流程與主觀權重影響結果的缺點，並可驗證本研究模型的可行性與有效性，在實務決策上將更適宜醫院建築維護主管應用與參考。

Large hospitals such as medical centers provide not only medical services, but also are responsible for emergency refuges, medical research and education. With the increasing demands on healthcare facilities and services, hospital buildings have gradually matured to become large capacity and numerous complex facilities. A hospital is a more complex system of interacting environments, in comparison with other kinds of buildings. The performance of hospital buildings depends highly upon the efficiency of maintenance execution.
However, with ever-growing demands and decreasing budgets, facilities managers of hospitals must ensure facilities properly maintained without compromising their performance. In order to provide an appropriate environment in such a high-complex building, key maintenance items should be carefully analyzed and executed as well as decisions for proper maintenance budget allocation. Therefore, the purposes of this study are to analyze the key maintenance items of hospital buildings, establish the prediction model for maintenance cost, and develop the decision model for alternatives priority.
In the aspect of key maintenance items analysis, three categories are classified, including periodic maintenance, repair, and demand change. This study also identifies the key maintenance items and their sub-items of the National Taiwan University Hospital (NTUH). Besides, this study also compares the differences between NTUH and literatures based on the maintenance cost unit. Results show that:
1. In periodic maintenance, the key maintenance items of both East building and West Building are High voltage (22.1% in East building and 26.1% in West Building), HVAC (26.2% in East building and 25.9% in West Building), and Elevator (13.9% in East building and 15.9% in West Building).
2. In repair, the key maintenance items of East building and West building are: (1) Interior finishes (43.4%), HVAC (24.7%), and Low voltage (9.3%) in East building; (2) Interior finishes (67.3%), Low voltage (10.6%), and HVAC (5.1%) in the West building. The average cost of West building is 2.8 times than East building (22.8 dollars/m2/year in West building and 8 dollars/m2/year in East building).
3. In demand change, the key maintenance items of East building and West building are: (1) Low voltage (52.7%), Boiler (31.3%), Interior finishes (11.1%), and HVAC (4.1%) in East building; (2) Interior finishes (29.9%), Low voltage (27.6%), HVAC (17.9%), and Elevator (16.0%) in the West building.
This study not only identifies the key maintenance items but also establishes an effective prediction model for hospital building maintenance, evaluating the tendency of maintenance cost to provide important information for the facility managers. In the aspect of prediction model, this study adopts the grey forecasting model as the main structure and then integrates the exponential series and self-organizing mapping to recognize the regular and irregular residual errors, respectively. The repair records of NTUH are employed to demonstrate the practicability and precision. Results show that the proposed prediction model can catch the tendency of building maintenance cost effectively.
Beside, this study provides a decision priority model for maintenance alternatives. The concepts of grey number, grey relational analysis, and weight-setting technique are all integrated into the proposed decision priority model. By defining the uncertain information, the proposed model can flexibly deal with the complex decision problems.

口試委員會審定書 Ⅰ
誌謝 Ⅱ
中文摘要 Ⅳ
英文摘要 Ⅵ
第一章 緒論 1
1.1 研究動機 1
1.2 研究目的 2
1.3 研究範圍與限制 3
1.4 研究流程與方法 3
第二章 文獻回顧 6
2.1 醫院建築發展趨勢 6
2.2 醫院建築維護管理 10
2.3 灰色系統理論 15
2.4 類神經網路 17
第三章 台大醫院建築維護成本分析與探討 23
3.1 概論 23
3.1.1 台大醫院建築簡介 23
3.1.2 資料蒐集方式 25
3.2 維護成本項目分析 33
3.2.1 東西址院區之定期維護成本 33
3.2.2 東西址院區之損壞修繕成本 36
3.2.3 東西址院區之需求變更成本 38
3.3 維護修繕單位成本分析與比較 39
3.4 維護修繕關鍵項目之策略分析 41
第四章 醫院建築維護成本預測模型建構與應用 44
4.1 概論 44
4.2 兩階段灰色殘差修正預測模型建構 44
4.2.1 灰色系統模型 45
4.2.2 指數級數殘差修正模型 49
4.2.3 自組織映射圖網路殘差修正模型 52
4.3 醫院維護成本預測案例 57
4.3.1 案例介紹 57
4.3.2 預測結果分析與比較 60
第五章 醫院建築維護決策模型建構與應用 67
5.1 概論 67
5.2 建築維護決策模型建構 67
5.3 醫院維護決策案例 77
第六章 結論與建議 83
6.1 結論 83
6.2 建議 85
參考文獻 88

國外文獻
1.Alippi, C., and Piuri,V., Experimental neural networks for prediction and identification. IEEE Transactions on Instrumentation and Measurement 45:670-676 (1996).
2.Baba, N., Kozaki, M., An intelligent forecasting system of stock price using neural networks. Proceedings of IJCNN 1:371-377 (1992).
3.Bingqian, L., The grey modelling for central symmetry sequence. The Journal of Grey System 2:95-103 (1990).
4.Chen, C.K., and Tien, T.L., The indirect measurement of tensile strength by the deterministic grey dynamic model DGDM(1, 1, 1). International Journal of Systems Science 28:683-690 (1997).
5.Chen, H.S., Residual updating of GM(1,1) model by exponential series. The Journal of Grey Systems 3:151-155 (2000).
6.Cheng, C.S., Hsu, Y.T., Wu, C.C., Grey modeling using neural network. The Journal of Grey System 3:219-232 (1997).
7.Chi, S.C., Chen, H.P., Chend, C.H., A forecasting approach for stock index future using grey theory and neural networks. Proceedings of the IJCNN 6:3850-3855 (1999).
8.Deng, J.L., Control problem of grey system, System and Control Letters 1:288-294 (1982).
9.Hsia, K.H., Chen, M.Y., and Chang, M.C., Comments on Data Pre-processing for Grey Relational Analysis, Journal of System, 7: 15-20 (2004).
10.Hsu, Y.T., Cheng, C.S., Wu, C.C., Grey modeling using neural network. The Journal of Grey System 9:219-232 (1997).
11.Hsu, Y.T., Lin, C.B., and Cheng, H.C., Grey model selections based on fuzzy trend prediction. Journal of Fuzzy Mathematics 7:897-905 (1999).
12.Jang, G.S., Lai, F.P., Wang, B.W., and Parng, T.M., Intelligent stock trading system with price trend prediction and reversal recognition using dual module neural network. Journal of Applied Intelligence 3:225-248 (1993).
13.Jang, J.S.R., Adaptive-network-based fuzzy inference systems. IEEE Transactions on Systems, Man, & Cybernetics 23:665-685 (1991).
14.Lavy, S., Shohet, I.M., Integrated maintenance management of hospital buildings: a case study, Construction Management and Economics 22:25-34 (2004).
15.Li, Z.L., New Definition of Correlation Function with Interval Number and Its Application in Extenics, Mathematics in practice and theory, 36(2): 207-210 (2006).
16.Lin, C.B., Su, S.F., and Hsu, Y.C., High precision forecast using grey models. International Journal of Systems Science 32:609-619 (2001).
17.Lin, C.T., Lee, C.S.G., Neural-network-based fuzzy logic control and decision systems. IEEE Transactions on Systems, Man, & Cybernetics 40:1414-1427 (1991).
18.Lin, Y.H., Lee, P.C., Novel high-precision grey forecasting model. Automation in Construction 16:771-777 (2007).
19.Lin, Y.H., Lee, P.C., Chang, T.P., and Ting, H.I, Multi-attribute group decision making model under the condition of uncertain information, Automation in Construction, 17: 792-797 (2008).
20.Liu, T.Y., Yeh, J., Chen, C.M., Hsu, Y.T., A grey ART system for grey information processing. Neurocomputing 56:407-414 (2004).
21.Luo, D., Liu, S.F., Study on the Method for Grey Incidence Decision-Making, Chinese Journal of Management Science, 13(1): 101-106 (2005).
22.Monin, Salut, G., ARMA lattice identification: a new hereditary algorithm. IEEE Transactions on Signal Processing 44:360-370 (1996).
23.Pan, C.L., Huang, Y.F., and Lin, G., Evaluation and Comparison of Data Pre-processing Transformer for Grey Relational analysis, Journal of Grey System 7: 65-74 (2004).
24.Shohet, I.M., Building evaluation methodology for setting maintenance priorities in hospital buildings, Construction Management and Economics 21:681-692 (2003).
25.Tan, C.L., Chang, S.P., Residual correction method of Fourier series to GM (1, 1) model. Proceedings of the First National Conference on Grey Theory and Applications, Kaohsiung, Taiwan 93-101 (1996).
26.Tan, C.L., Lu, B.F., Grey Markov chain forecasting model. Proceedings of the First National Conference on Grey Theory and Applications, Kaohsiung, Taiwan 157-162 (1996).
27.Tien, T.L., A research on the deterministic grey dynamic model with multiple inputs DGDMMI(1,1,1). Applied Mathematics and Computation 139:401-416 (2003).
28.Tseng, F.M., Tzeng, G.H., A fuzzy seasonal ARIMA model for forecasting. Fuzzy Sets and Systems 126:367-376 (2002).
29.Ture, M., Kurt, I., Comparison of four different time series methods to forecast hepatitis A virus infection. Expert Systems with Applications 31:41-46 (2006).
30.Uhlik, F.T., Hinze, J., Trends in the Construction Needs of Hospital Facilities, Journal of Architectural Engineering 4:132-134 (1998).
31.Valenzuela, H.M., Jabbi, A.S., 2-D recursive bilinear model for nonlinear signal. IEEE Transactions On Circuits and Systems II: Analog and Digital Signal Processing 40:241-250 (1993).
32.Wang, Z.X, Lu, C.G., and He, Y.H., An optimal method for multi-attribute decision-making with intervals, Journal of Huanggang Normal University, 25(3): 12-15 (2005).
33.Wu, J.H., Chen, C.B., An alternative form for grey relational grade, Journal of Grey System, 11: 7-12 (1999).
34.Wu, W.Y., Chen, S.P., A prediction method using the grey model GMC(1, n) combined with the grey relational analysis: a case study on Internet access population forecast. Applied Mathematics and Computation 169:198-217 (2005).
35.Yamaguchi, D., Li, G.D., Mizutani, K., Akabane, T., Nagai, M., and Kitaoka, M., On the Generalization of Grey Relational analysis, Journal of Grey system, 9(1): 23-34 (2006).
36.Yu, C.H., and Fan, Z.P., A Clustering Method Based on Interval Numbers for Multi-index Information, Journal of Northeastern University (Natural Science) 25(2): 183-186 (2004).
37.Yu, L.Y., Li, H.C., and Yu, Y.C., Possibility Degree Method for Group Multi-attribute Decision Making with Uncertainty, Journal of Shanghai University (Natural Science), 8(4): 353-356 (2002).
38.Zhang, J., Wu, D., and Olson, D. L., The Method of Grey Related Analysis to Multiple Attribute Decision Making Problems with Interval Numbers, Mathematical and Computer Modelling, 42: 991-998 (2005).
39.Zhang, J.J., Method of grey related analysis to multiple attribute decision making problem with interval numbers, Systems Engineering and Electronics, 27(6): 1030-1033 (2005).
40.Zhou, J., Fang, R., Li, Y., Zhang, Y., and Peng, B., Parameter optimization of nonlinear grey Bernoulli model using particle swarm optimization. Applied Mathematics and Computation 207:292-299 (2009).
41.IFMA, http://www.ifma.org/what_is_fm/fm_definitions.cfm.
 
國內文獻
1.王雅慧，大學實驗型教學校舍生命週期成本分析－著重營運維護管理階段，國立台灣大學碩士論文，2005。
2.何斌、蒙清，灰色預測模型拓廣方法研究，系統工程理論與實踐，9：138-141，2002。
3.李群，灰色預測模型的進一步拓廣，系統工程理論與實踐，13：64-66，1993。
4.李維倫，建築排水系統壽期評估方法與可視化模擬之研究，國立台灣科技大學碩士論文，2008。
5.李翠鳳、戴文戰，基於函數變換的灰色建模方法，系統工程，23：110-114，2005。
6.林明志，學校建築生命週期履歷表之研究 - 以台北市國民小學為例，國立台灣大學碩士論文，1997。
7.洪子茵，台北市集合住宅管理維護模式之研究，國立政治大學碩士論文，2002。
8.徐春福，公寓大廈共用部分機電設備維護費用支出之研究-以台中市為例，中華大學碩士論文，2002。
9.夏郭賢、吳漢雄，灰關聯分析之線性數據前處理探討，灰色系統學刊，1：47-53，1998。
10.夏郭賢、陳明儀，供應商評選模式分析－片段線性數據前處理法之應用，灰色系統理論與應用學術研討會，頁D1-D4，2001。
11.陳濤捷，灰色預測模型的一種拓廣，系統工程，8：50-52，1990。
12.陳燕菁，學校建築維護管理之探討-以台北市國民小學為例，國立台灣大學碩士論文，1997。
13.陳東興，醫院建築維護之案例式推理決策支援系統，國立臺灣科技大學碩士論文，2006。
14.張偉哲，廣義灰關聯生成模型之研究，灰色系統學刊，3：53-62，2000。
15.張智元，建築醫學之概念與應用機制研究，國立台灣大學博士論文，2005。
16.張家瑞，體育館建築生命週期成本之研究—以台大新舊體育館為例，國立台灣大學碩士論文，2005。
17.張佳瑜，智慧型建築維護管理成本探討，國立台北科技大學碩士論文，2008。
18.黃國勳，醫院工務室組織再造之研究-以某大型醫學中心工務室為例，國立台灣科技大學營建工程系碩士論文，2001。
19.溫坤禮、黃宜豊、張偉哲、張廷政、游美利、賴家瑞，灰關聯模型方法與應用，高立圖書有限公司，2003。
20.劉建志，台南市立國民中小學學校建築永續利用之研究－以修繕更新工程為主軸，國立成功大學碩士論文，2004。
21.戴佑安，以生命週期成本觀點論大學圖書館建築營運管理之研究，國立臺灣大學碩士論文，2007。
22.羅昱盈，以生命週期成本觀點研析學生宿舍建築之維護成本與策略，國立台灣大學碩士論文，2005。