由於經濟發展迅速，都會區人口密集，為疏解人潮及解決交通問題，新建大眾捷運系統成為必要的趨勢，並帶動捷運沿線快速開發及促進地方繁榮，因此捷運沿線的新建案件大量增加；但臺北盆地松山層由鬆軟泥砂互層組成，先天地質環境條件不佳，鄰近施工對潛盾隧道損壞風險相對提高，近期以臺北大巨蛋鄰近板南線施工造成潛盾隧道徑向變位超出行動值及部分軌道面傾斜，使臺北市捷運局與建商之間紛擾不斷成為新聞焦點，引起社會不安，故鄰近施工對潛盾隧道變形的影響可成為深入探討的議題。
目前工程界常用以FLAC及PLAXIS等數值分析軟體應用於鄰近施工影響潛盾隧道變形的分析，為模擬真實工程狀況，且對不同界面模擬相互作用，需要有經驗的專任人員加入與過往類似實際案例回饋資料，才能保證分析結果的合理性，因此通常需由專業人員操作需求、建模複雜度及難度較高等問題。
本研究嘗試以人工智慧推論模式對鄰近施工引致捷運潛盾隧道徑向變位建立一套預測模式，並透過文獻及SPSS相關性驗證，確認潛盾隧道徑向變位影響因子，建立案例資料庫，本研究利用「NNLSTM」案例訓練與測試後，結果顯示「NNLSTM」的誤差衡量指標優於其他推論模式，擁有最佳的預測能力，後續將以「NNLSTM」做為捷運潛盾隧道徑向變位推論模式。
本研究應用「NNLSTM」於實例中對6處觀測點預測最大潛盾隧道徑向變位與實際值比較，並利用MAPE做誤差衡量指標，研究結果顯示CP4、CP6、CP7、CP9、CP11等5處預測誤差率在10 %以內，屬於精準預測，另CP8觀測點預測誤差率14.01 %，屬於優良預測，分析其原因為NNLSTM可分別對非時序性因子及時序性因子進行演算，並將兩模式輸出結果整合，得到最後預測值並結合自動調整權重功能，提高實際預測精準度。
因此「NNLSTM」應用於潛盾隧道徑向變位預測有使用簡便性、高精確準度及可靠度高等優點。可應用本研究的推論模式來預測各階段的潛盾隧道徑向變位，協助施工管理人員提早發現異常情形及預警參考，以採取適當的處理措施，做有效的風險管控。

Due to the rapid economic development and dense population in the metropolitan area, the new mass transit system has become a necessary trend, and the rapid development of the MRT and the promotion of local prosperity have led to a large increase in new cases along the MRT; However, the Songshan Formation in the Taipei Basin is composed of soft mud-sand interbeds. The congenital geological environment conditions are not good. The risk of damage to the shield tunnels is relatively high in the adjacent construction. Recently, the radial displacement of the shield tunnels caused by the construction of the Taipei Big Giant near the Bannan Line has exceeded the action. The value and the inclination of part of the orbital surface have caused the disturbance between the Taipei City MRT and the construction company to become the focus of the news and cause social unrest. Therefore, the influence of the adjacent construction on the deformation of the shield tunnel can be an in-depth discussion.
At present, the engineering community often uses the numerical analysis software such as FLAC and PLAXIS to analyze the deformation of the shield tunnel affected by adjacent construction. In order to simulate the real engineering situation and simulate the interaction of different interfaces, it is necessary for experienced full-time personnel to join the actual case. Feedback data can ensure the rationality of the analysis results, so it is usually necessary for professionals to operate the requirements, modeling complexity and difficulty.
This study attempts to establish a set of prediction modes for the radial displacement of the MRT submarine tunnel caused by adjacent construction by artificial intelligence inference model. Through the literature and SPSS correlation verification, confirm the radial displacement influence factor of the shield tunnel and establish case data. In this study, after using the "NNLSTM" case training and testing, the results show that the error measure of "NNLSTM" is superior to other inference models and has the best predictive ability. The follow-up will use "NNLSTM" as the MRT tunnel path. Inferion mode to displacement.
In this study, we use "NNLSTM" to predict the maximum displacement of the maximum shield tunnel from the six observation points and compare the actual values, and use MAPE as the error measure. The research results show that CP4, CP6, CP7, CP9, CP11, etc. 5 The prediction error rate is within 10%, which is accurate prediction. The prediction error rate of CP8 observation point is 14.01%, which is a good prediction. The reason is that NNLSTM can calculate the non-sequential factors and time series factors separately, and the two modes. The output results are integrated, and the final predicted value is combined with the automatic adjustment weight function to improve the actual prediction accuracy. Therefore, "NNLSTM" is applied to the prediction of radial displacement of the shield tunnel, which has the advantages of ease of use, high precision and high reliability. The inference model of this study can be used to predict the radial displacement of the shield tunnel at each stage, and assist the construction management personnel to detect the abnormal situation and early warning reference early, so as to take appropriate treatment measures and effectively control the risk.

[1]陳伯興、余旭洲，2014，潛盾隧道受鄰近工區近接施工之影響評估及分析，捷運技術半年刊，第48期，第213-234頁
[2]胡邵敏，2015，明日的深開挖工程~由臺北大巨蛋談起，地工技術，第145期，第97~101頁
[3]高宗正，陳俊宏，李維峰，2015，近接施工規範以及審查制度推動成果與未來展望，地工技術，第144期，第7~14頁
[4]交通部、內政部，2012，大眾捷運系統兩側禁建限建辦法
[5]臺北市政府，2016，臺北市辦理臺北都會區大眾捷運系統禁限建範圍內列管案件管理及審核基準
[6]黃南輝，2011，地下工程與風險管理，臺灣省及臺北市大地技師公會
[7]歐晉德、李延恭、鄭在仁，1983，臺北盆地松山層地下水位及水壓分佈對基礎工程影響，土木水利，第10卷，第3期，第89~102頁
[8]吳特偉，1988，臺北盆地地盤分區土壤之工程特性，地工技術，第22期，第5~27頁
[9]陳俊宏、段紹緯、黃南輝，2002，臺北捷運潛盾施工引致沉陷探討，地工技術交流研討會論文集
[10]Palmer J H L V, Kenney T C. Analytical study of a braced excavation in weak clay[J]. Canadian Geotechnical Journal, 1972, 9(2): 145-164
[11]Mana , A.I. and Clough , G.W.,1981,Prediction of Movement for Braced Cut in Cay , Journal of the Geotechnical Division , ASCE , Vol.117,No.6,
pp 1590-1609
[12]Clough,G.W. and O'Rourke,T.D.,Construction induced movements of insitu walls,Proceedings,Design and performance of earth retaining structures, ASCE, Special Conference,Ithaca,New York,1990: 439-470.
[13]Wong , K.S. and Broms , B.B.,1989,Lateral Wall Deflections of Braced Excavation in Weak Clay , Journal of Geotechnical Engineering , Vol.115, No.6, PP.853-870
[14]Masuda , T. , Einstein,H.H. and Mitachi , T.,1994 , Prediction of Lateral Deflection of Diaphragm Wall in Deep Excavation , Jounal of Geotechnical Engineering , Proceeding of Japan Society of Civil Engineers,NO.505,ш-29 ,PP.19-29.
[15]陳煌銘，1987，深開挖引致之地盤移動，地工技術，第 20 期，第19-33頁
[16]平岡成明、石正義，1987， 假設工程問題對策 ，詹氏書局，第 20 期，第19-33頁
[17]歐章煜，2002，深開挖工程分析設計理論與實務，科技圖書股份有限公司
[18]黃南輝、張培義、王復國、簡敏忠，2000，臺北捷運深開挖經驗，科技圖書
[19]林雅雯，1999，地中壁應用於深開挖之分析研究，國立臺北科技大學，碩士論文
[20]鄭光宏，2001，內扶壁對連續壁側向變位之影響，國立雲林科技大學，碩士論文
[21]簡文青，2002，扶壁改善深開挖擋土壁體變形行為之研究，國立中央大學，碩士論文
[22]李宗霖，2005，內扶壁在基一區(K1)對連續壁變位之影響，國立中央大學，碩士論文
[23]謝百鈎，林亦郎，黃國樑，2006，應用三項度分析探討地中壁用於深開挖之分析與設計，行政院國家科學委員會專題研究計畫
[24]盧怡志，2007，影響扶壁及地中壁行為施工特性之研究，國立臺北科技大學，碩士論文
[25]李崇道，2009，深開挖之扶壁數值分析，國立臺北科技大學，碩士論文
[26]鄧文賓，2013，扶壁對深開挖壁體變形影響之研究，國立臺灣科技大學，碩士論文
[27]謝旭昇，2015，地改,扶壁及地中壁之配置原則，簡報
[28]黃之辰，2015，深開挖中扶壁及地中壁在不同影響因子下之效率分析，國立臺北科技大學，碩士論文
[29]王繼勝、李耀明，1993，深開挖工程對潛盾隧道應力之影響，結構工程，第8卷，第3期，第95~103頁
[30]劉玫瑛，1994，深開挖與鄰近地下結構互制行為之三維分析，國立中央大學，碩士論文
[31]余明山、賴慶和、楊清源，1996，黏土層終身開挖對鄰近潛盾隧道影響分析，地工技術，第53期，第67~74頁
[32]陳鴻濤、彭嚴儒，1997，深開挖工程對鄰近隧道之影響，營建知訊，第178期，第16~23頁
[33]王建智、林宏達、冀樹勇，2001，深開挖對鄰近潛盾隧道影響之分析，行政院國家科學委員會補助專題研究計畫
[34]何泰源、陳卓然、陳福勝，2003，潛盾隧道襯砌環片與土壤互制行為之探討，地工技術，第97期，第77~86頁
[35]王建智、詹勳山、謝坤宏，2003，深開挖對鄰近現有潛盾隧道之變形分析，技術學刊，第18卷，第2期，第235~243頁
[36]馮正一、陳德郎，2007，深開挖對臨近捷運結構物之影響，第12屆大地工程研討會
[37]王廷瑋、陳俊融、廖惠生，2015，潛盾隧道上方開挖之影響及對策，地工技術，第144期，第57~64頁
[38]M.-Y. Cheng , A.F.V. Roy, 2011, Evolutionary fuzzy decision model for cash flow prediction using time-dependent support vector machines. International Journal of Project Management, 29(1): 56-65.
[39]M.-Y. Cheng , Y.-W. Wu, 2009, Evolutionary support vector machine inference system for construction management, Automation in Construction, 18(5): 597-604.
[40]M.-Y. Cheng, D. Prayogo, Y.-W. Wu, 2014, Novel Genetic Algorithm-Based Evolutionary Support Vector Machine for Optimizing High-Performance Concrete Mixture. Journal of Computing in Civil Engineering, 28(4): 06014003.
[41]M.-Y. Cheng, M.-T. Cao, 2014, Evolutionary multivariate adaptive regression splines for estimating shear strength in reinforced-concrete deep beams. Engineering Applications of Artificial Intelligence, 28: 86-96.
[42]D. Prayogo, 2012, A Novel Genetic Algorithm-Based Evolutionary Support Vector Machine for Optimizing High-Performance Concrete, Master Thesis, Department of Construction Engineering, National Taiwan University of Science and Technology.
[43]M.-Y. Cheng, N.-D. Hoang, Y.-W. Wu, 2013, Hybrid intelligence approach based on LS-SVM and Differential Evolution for construction cost index estimation: A Taiwan case study. Automation in Construction, 35: 306-313.
[44]M.-Y. Cheng, P.M. Firdausi, D. Prayogo, 2014, High-performance concrete compressive strength prediction using Genetic Weighted Pyramid Operation Tree (GWPOT). Engineering Applications of Artificial Intelligence, 29: 104-113.
[45]M.-Y. Cheng, D.K. Wibowo, D. Prayogo, A.F.V. Roy, 2015, Predicting productivity loss caused by change orders using the evolutionary fuzzy support vector machine inference model. Journal of Civil Engineering and Management, 21(7): 881-892.
[46]D.E. Goldberg, K. Deb, H. Kaegupta, and G. Harik, 1993, Rapid, accurate optimization of difficult problems using fast messy genetic algorithms.
[47]Suykens, J., et al., 2002, Least Square Support Vector Machines. World Scientific Publishing Co. Pte. Ltd.
[48]Min-Yuan Cheng and Nhat-Duc Hoang, 2012, Evolutionary Least Squares Support Vector Machine – Userguide, Technical Report, CIC Lab, National Taiwan Univ. of Sci. and Tech.
[49]Min-Yuan Cheng and Nhat-Duc Hoang, “Estimating Compressive Strength of Rubberized Concrete Using Evolutionary Least Squares Support Vector Machine”,Technical Report, CIC Lab, National Taiwan Univ. of Sci. and Tech,2012.
[50]Min-Yuan Cheng and Doddy Prayogo, “Symbiotic Organisms Search: A new metaheuristic optimization algorithm”, Computers & Structures, Vol. 139, pp. 98-112, 2014.
[51]G.G. Tejani, V.J. Savsani, Patel V. K., 2016, Adaptive symbiotic organisms search (SOS) algorithm for structural design optimization. Journal of Computational Design and Engineering.
[52]M.-Y. Cheng, D. Prayogo, D.-H. Tran, 2015, Optimizing Multiple-Resources Leveling in Multiple Projects Using Discrete Symbiotic Organisms Search. Journal of Computing in Civil Engineering: 04015036.
[53]D.-H. Tran, M.-Y. Cheng, D. Prayogo, 2016, A novel Multiple Objective Symbiotic Organisms Search (MOSOS) for time–cost–labor utilization tradeoff problem. Knowledge-Based Systems, 94: 132-145.
[54]M.-Y. Cheng, C.-K. Chiu, Y.-F. Chiu, Y.-W. Wu, Z.-L. Syu, D. Prayogo, C.-H. Lin, 2014, SOS optimization model for bridge life cycle risk evaluation and maintenance strategies. Journal of the Chinese Institute of Civil and Hydraulic Engineering, 26(4): 293-308.
[55]S. Duman, 2016, Symbiotic organisms search algorithm for optimal power flow problem based on valve-point effect and prohibited zones. Neural Computing and Applications: 1-15.
[56]H. Kamankesh, V.G. Agelidis, A. Kavousi-Fard, 2016, Optimal scheduling of renewable micro-grids considering plug-in hybrid electric vehicle charging demand. Energy, 100: 285-297.
[57]E. Ruskartina, V.F. Yu, B. Santosa, A.A.N.P. Redi, 2015, Symbiotic Organism Search (SOS) for Solving the Capacitated Vehicle Routing Problem. International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering, 101: 857-861.
[58]A. Panda, S. Pani, 2016, A Symbiotic Organisms Search algorithm with adaptive penalty function to solve multi-objective constrained optimization problems. Applied Soft Computing, 46: 344-360.
[59]Min-Yuan Cheng, Doddy Prayogo and Yu-Wei Wu, (in preparation), “Predicting the Pavement Rutting Behavior of Asphalt Mixtures Using Symbiotic Organisms Search - Least Squares Support Vector Machine Inference Model”, Construction and Building Materials.
[60]Sepp Hochreiter and Jurgen Schmidhuber(1997), “Long short-term memory. Neural Computation,” 9(8):1735-1780.
[61]Felix A. Gers, Jurgen Schmidhuber, Fred Cummins.(2000), “Learning to forget:Continual prediction with LSTM. Neural computation”, 12(10):2451-2471
[62]Ilya Sutskever, Oriol Vinyals, and Quoc V. Le.(2014), “Sequence to sequence learningwith neural networks.”, In Advances in Neural Information Processing Systems,pages 3104-3112
[63]Lewis, C. D., 1982, International and Business Forecasting Methods. London: Butterwo
[64]應亦凡，2017，深開挖變形自調適動態學習預測模式之研究應用NNLSTM，國立臺灣科技大學，碩士論文
[65]安全監測計畫書，2014，冠德建設股份有限公司
[66]廖惠生、賴慶和、陳俊宏，2015，連續壁近接施工引致捷運潛盾隧道受損案例之探討，地工技術，第144期，第47~56頁