準確預估船舶成本是船廠的核心能力，它可增加獲得船東訂單的機會。現有初步階段的成本預估方法，利用數個船舶主要參數，來預估人工、材料與裝備的成本，但因未能充分考慮到設計、材料、施工、匯率與市場等成本變動的因素，導致誤差較大。因此本研究之主旨為發展船舶主要成本項目的預估方法，將船舶鋼料、主機與發電機及建造人工成本等重要因素納入研究。
本文提出主要特徵預估法，建立主要尺寸參數與細部特徵的關聯性，來預估各項船舶規格。首先就船舶成本相關的資訊，探勘參數及細部設計與成本項目的關係，選取重要特徵，再轉化這些特徵為量化的型態。其次定義主要特徵與次要特徵，主要特徵一般是船舶主要參數的組合，採用數個關鍵性的參數，代表預估項目的核心成分。利用主要特徵參數結合一到兩個次要特徵參數，經迴歸分析，預估樣本的主要趨勢。然後使用次要特徵修正前述預估的誤差，特別考慮了船殼三維幾何變化、主要施工、材料與裝備導致的因素，由組合預估參數或加入經驗參數，對應誤差的原因，來修正預估結果。
本文預估貨櫃船的鋼料重量，先對船體結構區分艙區，決定預估的特徵項目，包含船艏、貨艙、機艙、船艉及住艙等分段。由主要尺寸與各分段的長度組合，對應受船體線型影響的縱向單位長度重量變化，考慮各船段越近船艏位置的重量越輕，越近機艙位置的重量越重的特性。由預估各分段鋼料重量可得知加總的全船鋼料重量，對各船型的平均誤差約1.15%。與現有其它四種方法比較的平均誤差分別是3.36%、3.97%、4.39%與5.37%。可見主要特徵預估法較為準確。接著，透過比較近期兩型貨櫃船，運用研究發展的各項預估方程式，預估重要材料與裝備的誤差約在±8%。最後，計算全船成本誤差分別為-2.63%與0.84%。因此說明具備初步概算成本的價值，後續可作為其它船型成本預估方法的研究基礎。

Accurate cost estimation is crucial for obtaining ship owners’ orders in shipyards. The preliminary classic estimation methods of ship costs, only provide rough estimates of the labor, material and equipment based on overall ship parameters, do not reflect further specifications, including design, material, construction, exchange rate, and marketing costs. In this study, an innovative cost estimation method, named principal feature estimation method, is developed based on the preliminary specifications for estimating ship costs, including the steel and other main material, engine, power generator and other core equipment, man-hours, etc.
The method mainly establishes the topology of the relationships of the feature by linking the principal dimension parameters and detailed features of the specifications of designs and cost information for estimating the main specifications of ship. First, the parameters and relations of the design and cost of ships are mined for the principal and secondary features based the design and cost analysis. Then extract features and transforms them into a quantifiable structure. Second, the definition of features consists the principal and secondary features. The principal features, which are general parameters using the parameter combinations, contain the core context using a little information. The estimation based on the combination parameters of principal and secondary features could account for the approximate most composition of the actual value. Third, the estimate may be adjusted for various characteristics, that are neglected, based on the secondary features using the special treatments, especially three-dimensional hull form design, manufacturing style, material and equipment specifications.
The design of a ship hull is based on the specific geometric definition of a hull which influences the hydrostatics, general arrangement, strength and aesthetics of the vessel. A naval architect has to apply his combined knowledge of designing and draftsmanship to arrive at the optimum hull form. An important aspect of preliminary design is to create a set of faired ship lines.
First, the method for estimating the steel weight of containerships is based on the features of structural segment, including the fore, cargo hold, engine room, aft, and deckhouse, etc. For improving hydrodynamic performance, the sectional area, below the waterline and near fore or aft peak, may be gradually reduced based on the fairness of the optimum hull form. It causes the weight to be concentrated in the aft portion of the segment. And the segment becomes lighter as the segment becomes closer to the bow. Several methods are used to compare the estimation variability: the average error for the presented method, which sums the weights for each section, is about 1.15%. Meanwhile, the corresponding errors for the traditional methods are 3.36%, 3.97, 4.39, and 5.37%. The study finds that the principal feature estimation method clearly provides a lower average error and a smaller error range. Thus, this method is able to provide a more accurate steel weight.
Next, two recently built ships are examples for estimating the cost, including the items of the major material and equipment according to the developed equations. The total errors of the estimated material and equipment costs are less ±8%. Then the errors of the estimated total costs are -2.63% and 0.84%. Hence, the estimated model is suitable for these modern target ships. The applications of the model may be more robust for new ships in the future study.

1. Aasen, R., Bjorhovde, S. (2010), “Early Stage Weight and Cog Estimation
Using Parametric Formulas and Regression on Historical Data”, 69th Annual Conference, Society of Allied Weight Engineers, Inc., Virginia Beach, Virginia, pp. 1-35.
2. ABS (2011), Rules for Building and Classing Steel Vessels, American Bureau of Shipping.
3. Atchley, W.R., Bryant, E.H. (1975), Multivariate statistical methods, among-groups covariation, Halsted Press.
4. Barentine, J.M. (1996), A process-based cost estimating tool for ship structural designs, Department of Ocean Engineering, Naval Postgraduate School, Monterey, California.
5. Barrass, B. (2004), Ship design and performance for masters and mates, Butterworth-Heinemann.
6. Benford, H. (1967), “The practical application of economics to merchant ship design”, Journal of Marine Science Technology, 1 (1) pp. 519-536.
7. Bertram, V., Schneekluth, H. (1998), Ship design for efficiency and economy, Butterworth-Heinemann.
8. Beveridge, J.L. (1971), “Design and performance of bow thrusters”. DTIC Document.
9. Bole, M. (2007), “Cost assessment at concept stage design using parametrically generated production product models”, International Conference on Computer Applications in Shipbuilding, pp. 1-13.
10. Caprace, J.-D., Rigo, P. (2012), “Towards a short time feature-based costing for ship design”, Journal of Marine Science and Technology, 17 (2) pp. 216-230.
11. Caprace, J.D., Rigo, P. (2011), “Ship complexity assessment at the concept design stage”, Journal of Marine Science and Technology, 16 (1) pp. 68-75.
12. Carreyette, J. (1978), “Preliminary ship cost estimation”, Naval Architect, 1 (4) pp. 235-258.
13. Celik, F., Erturk, I., Turan, E. (2013), “Investigation of main particulars subject to minimum building cost for chemical tankers”, Ocean Engineering, 73 (1) pp. 32-37.
14. Christensen, W.L., Chirillo, L.D., Maguire, S., Gambello, A. (1992), Self Assessment of Advanced Shipbuilding Technology Implementation, Defense Technical Information Center.
15. DNV (2011), Strength Analysis of Hull Structures in Container Ships, Det Norske Veritas.
16. Fischer, J.O., Holbach, G. (2011), Cost management in shipbuilding, GKP Publishing, Koeln.
17. Geiger, T.S., Dilts, D.M. (1996), “Automated design-to-cost: integrating costing into the design decision”, Computer-Aided Design, 28 (6) pp. 423-438.
18. Hart, C.G., He, Z., Sbragio, R., Vlahopoulos, N. (2012), “An advanced cost estimation methodology for engineering systems”, Systems Engineering, 15 (1) pp. 28-40.
19. Hart, C.G., Vlahopoulos, N. (2010), “An integrated multidisciplinary particle swarm optimization approach to conceptual ship design”, Structural and Multidisciplinary Optimization, 41 (3) pp. 481-494.
20. Hawkins, S., Taggart, R., Hoyt, E.D. (1965), “The use of maneuvering propulsion devices on merchant ships”.
21. Hotelling, H. (1933), “Analysis of a complex of statistical variables into principal components”, Journal of Educational Psychology, 24 (6) pp. 417-441.
22. It, J. (2002), Principal component analysis, New York, Springer Series in Statistics.
23. Jackson, J.E. (2005), A user's guide to principal components, John Wiley & Sons.
24. Joshi, B.K., Panthee, D.R. (2011), “Information Theory and Multivariate Techniques for Analyzing DNA Sequence Data: An Example from Tomato Genes”, Nepal Journal of Biotechnology, 1 (1) pp. 1-9.
25. Kaluzny, B.L., Barbici, S., Berg, G., Chiomento, R., Derpanis, D., Jonsson, U., Shaw, R.D., Smit, M.C., Ramaroson, F. (2011), “An application of data mining algorithms for shipbuilding cost estimation”, Journal of Cost Analysis and Parametrics, 4 (1) pp. 2-30.
26. Kazuhiko, M. (1998), “Rough estimation of light weight and ship price”, Science of Ships, 51 (12) pp. 59-63.
27. Kerlen, H. (1985), Über den Einfluss der Völligkeit auf die Rumpfstahlkosten von Frachtschiffen, Inst. für Schiffbau.
28. Kwon, S., Kim, B.C., Mun, D., Han, S. (2015), “Simplification of feature-based 3D CAD assembly data of ship and offshore equipment using quantitative evaluation metrics”, Computer-Aided Design, 59 (1) pp. 140-154.
29. Løseth, R., Sekkesæter, G., Valsgård, S. (1994), “Economics of high-tensile steel in ship hulls”, Marine Structures, 7 (1) pp. 31-50.
30. Lamb, T. (2004), Ship design and construction, The Society of Naval Architects and Marine Engineers (SNAME), New Jersey.
31. Lee, S.H. (2005), “A CAD–CAE integration approach using feature-based multi-resolution and multi-abstraction modelling techniques”, Computer-Aided Design, 37 (9) pp. 941-955.
32. Lin, C.-K., Shaw, H.-J. (2015), “Feature-based estimation of steel weight in shipbuilding”, Ocean Engineering, 107 (1) pp. 193-203.
33. Lyon, T.D., Mistree, F. (1985), “A computer-based method for the preliminary design of ships”, Journal of Ship Research, 29 (4) pp. 251-269.
34. Mano, Shigemasa (1991), “Memorandum on hull structural design”, Science of Ships, 44 (1) pp. 99-104.
35. Miller, D.S. (1968), “The Economics of the Container Ship Subsystem”. University of Michigan.
36. Molland, A.F., Turnock, S.R., Hudson, D.A. (2011), Ship resistance and propulsion: practical estimation of propulsive power, Cambridge university press.
37. Mulligan, R.F. (2008), “A simple model for estimating newbuilding costs”, Maritime Economics & Logistics, 10 (3) pp. 310-321.
38. NATO (2006), Ship Costing, North Atlantic Treaty Organization.
39. Nowacki, H. (2010), “Five decades of Computer-Aided Ship Design”, Computer-Aided Design, 42 (11) pp. 956-969.
40. Papanikolaou, A. (2014), Ship Design: Methodologies of Preliminary Design, Springer.
41. Pearson, K. (1901), “On lines and planes of closest fit to systems of points in space”, The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 2 (11) pp. 559-572.
42. Ross, J.M. (2004), “A practical approach for ship construction cost estimating”, Conference on Computer Applications and Information Technology in the Maritime Industries, Siguenza, Spain, pp. 98-110.
43. Sato, S. (1967), Effects of Principal Dimensions on Weight and Cost of Large Ships, Society of Naval Architects and Marine Engineers, New York
44. Shetelig, H. (2013), Shipbuilding Cost Estimation: Parametric Approach, Department of Marine Technology, Norwegian University of Science and Technology, Trondheim.
45. Son, M.-J., Lee, S.C., Kwon, K.-C., Kim, T.-W., Sharma, R. (2011), “Configuration estimation method for preliminary cost of ships based on engineering bills of materials”, Journal of Marine Science and Technology, 16 (4) pp. 367-378.
46. Turbo, M.D. (2011), Basic Principles of Ship Propulsion, MAN Diesel & Turbo, Munich.
47. Watson, D., Gilfillan, A. (1977), “Some ship design methods”, Naval Architect, 1 (4) pp.
48. Watson, D.G.M. (2002), Practical Ship Design, Elsevier.
49. Woodyard, D. (2009), Pounder's marine diesel engines and gas turbines, Butterworth-Heinemann.
50. Yang, Y.-S., Park, C.-K., Lee, K.-H., Suh, J.-C. (2007), “A study on the preliminary ship design method using deterministic approach and probabilistic approach including hull form”, Structural and Multidisciplinary Optimization, 33 (6) pp. 529-539.
51. Zakki, A.F. (2013), “The Assessment of CSR Regulations Implementation on the Midship Strength and Structural Weight of 77.500 DWT Bulk Carrier”, The Indonesian Journal of Naval Architecture, 1 (1) pp. 1-7.
52. Zhang, P., Zhu, D.-x., Leng, W.-h. (2008), “Parametric approach to design of hull forms”, Journal of Hydrodynamics, Ser. B, 20 (6) pp. 804-810.
53. 造船テキスト研究会 (2009), 商船設計の基礎知識, 成山堂書店.