本研究以紡織業織布廠前準備作業為研究對象，探討製程參數對胚布在製品庫存量之影響，並以滿足訂單需求之前提下降低在製品庫存量為目標。由於織布廠在少量多樣與接單式生產環境下，若參數設定不當，易造成過量生產與庫存累積，因此本研究延續既有用紗試算邏輯，重建用紗試算系統，並結合田口方法（Taguchi Method）與反應曲面法（ResponseSurface Methodology, RSM），建立製程參數最佳化之決策支援流程。
本研究首先透過顯著因子篩選，確認紗粒重量（Cone weight）、單軸條數（End count）、丹尼數（Denier）與軸併數（Beam）為主要影響因子，並依文獻與實務條件重新設定參數值域。接著，針對不同需求區間進行田口方法分析，以 S/N（Signal-to-Noise Ratio）比與因子效果判定較佳參數組合；再以田口結果作為反應曲面法起始點，透過一階模式配適（First-Order Model Fitting）、最陡下降法（Steepest Descent Method）、中央合成設計（CentralComposite Design, CCD）與二階模式最佳化求得較佳解。此外，本研究新增需求區間四與需求區間五，以提升模型於低、高需求量情境下之適用性。
實證結果顯示，擴充參數值域後，田口方法與反應曲面法皆能降低多數需求區間之胚布平均在製品庫存量，其中反應曲面法可透過局部建模與限制條件最佳化取得更低庫存。納入軸併數後，部分需求區間仍可進一步改善。整體而言，本研究流程可協助織布廠於不同需求情境下快速取得較佳參數組合，降低過量生產風險，並提升參數設定之決策品質。

This study focuses on the preparation process of a textile weaving mill and investigates theeffects of process parameter combinations on greige fabric work-in-process inventory. The mainobjective is to reduce inventory while ensuring that production output satisfies order demand.In a high-mix, low-volume, and make-to-order production environment, mismatches betweenyarn preparation settings and actual order requirements may lead to overproduction andinventory accumulation. Therefore, this study extends the yarn usage calculation logic ofprevious research by reconstructing the original Excel VBA-based calculation system on theGoogle Colab platform, which serves as the basis for subsequent experimental design andoptimization analysis.
First, significant factor screening is conducted to identify cone weight, ends per warp beam,number of beams combined, and denier as the main factors affecting greige fabric work-in-process inventory. The feasible ranges of these parameters are then redefined based on literature,equipment specifications, and practical production conditions. Next, the Taguchi method isapplied to different demand ranges and inventory scenarios. The smaller-the-better signal-to-noise ratio and factor effect analysis are used to determine better parameter-level combinations.
The results obtained from the Taguchi method are then used as the initial points for ResponseSurface Methodology. Through first-order model fitting, the steepest descent method, face-centered central composite design, and second-order modeling, improved parametercombinations are identified within the constrained design region. In addition, low-demand andhigh-demand scenarios are included to enhance the applicability of the proposed model underdifferent order scales.
The empirical results show that, after expanding the parameter ranges, both the Taguchi methodand Response Surface Methodology can reduce the average greige fabric work-in-processinventory in most demand ranges. Compared with the Taguchi method, Response SurfaceMethodology further improves inventory reduction through local modeling and constrainedoptimization. After incorporating the number of beams combined as an additional control factor,several demand ranges achieve further inventory reduction, indicating that this factor providesadditional support for output adjustment and inventory improvement. Overall, the proposedintegrated procedure, combining the yarn calculation system, the Taguchi method, and ResponseSurface Methodology, can help weaving mills quickly identify better process parametercombinations under different demand scenarios, reduce the risk of overproduction, and improvethe decision quality of process parameter settings.

高育淳（2023）。最小化織布製程在製品庫存之最佳化製程參數之研究，國立成功大學製造資訊與系統研究所碩士論文，台南。
曾珞萍（2000）。以反應曲面法尋找多目標模擬模式之最佳解－以半導體封裝廠印字區為例，國立成功大學製造資訊與系統研究所碩士論文，台南。
經濟部統計處（2025）。台灣紡織及成衣工業產值統計資料。紡拓會整理。
Abd, A. A., Shabbani, H. J. K., Helwani, Z., & Othman, M. R. (2022). Experimental study and static numerical optimization of scalable design of non-adiabatic and non-isothermal pressure swing adsorption for biogas upgrading. Energy, 257, 124781.
Ahmed, F., Mills, A. C., Kwon, C. H., et al. (2025). Impact of weaving parameters on the performance of yarn-based interconnects in woven e-textiles: A comprehensive review. Journal of Engineered Fibers and Fabrics, 20, 1–22.
Asghar, A., Abdul Raman, A. A., & Wan Daud, W. M. A. (2014). A comparison of central composite design and Taguchi method for optimizing Fenton process. The Scientific World Journal, 2014, Article 869120.
Box, G. E. P., & Hunter, J. S. (1957). Multi-factor experimental designs for exploring response surfaces. The Annals of Mathematical Statistics, 28(1), 195–241.
Borkowski, J. J. (1995). Spherical prediction-variance properties of central composite and Box-Behnken designs. Technometrics, 37(4), 399–410.
Cotton Works. (2021). Weaving 101. Cotton Incorporated. Available: https://www.cottonworks.com/wp-content/uploads/2021/04/Weaving-101.pdf （2025/10/11取得）
Ferro, R., Cordeiro, G. A., Ordóñez, R. E. C., & Shiki, S. B. (2021). An optimization tool for production planning: A case study in a textile industry. Applied Sciences, 11(18), 8312.
Giovannitti-Jensen, A., & Myers, R. H. (1989). Graphical assessment of the prediction capability of response surface designs. Technometrics, 31(2), 159–171.
Hill, W. J., & Hunter, W. G. (1966). A review of response surface methodology: A literature survey. Technometrics, 8(4), 571–590.
Hisam, M. W., Dar, A. A., Elrasheed, M. O., Khan, M. S., Gera, R., & Azad, I. (2024). The versatility of the Taguchi method: Optimizing experiments across diverse disciplines. Journal of Statistical Theory and Applications, 23(4), 365–389.
Hossain, I., Alim, M. A., Uddin, N., Saha, S. S., Parvez, M. S., Maniruzzaman, M., Haque, M. S., & Mallick, A. (2025). Development and optimisation of green flame retardant properties of cotton fabric treated with chitosan nanoparticles and phosphate using Taguchi method. Scientific Reports, 15, Article 44257.
Kalayci, E., & Avinc, O. (2024). The disperse dyeing of polyetherimide fibers using orthogonal array (Taguchi) design. Fibers and Polymers, 25, 3045–3060.
Montgomery, D. C. (2020). Design and analysis of experiments (10th ed.). John Wiley & Sons.
Myers, R. H., Montgomery, D. C., & Anderson-Cook, C. M. (2016). Response surface methodology: Process and product optimization using designed experiments (4th ed.). John Wiley & Sons.
National Institute of Standards and Technology. (n.d.). Single response: Optimization subject to experimental region constraints. In NIST/SEMATECH e-Handbook of Statistical Methods.
National Institute of Standards and Technology. (2024). Taguchi designs. In NIST/SEMATECH e-Handbook of Statistical Methods.
Önder, E., & Berkalp, Ö. (2017). Weaving technology II: Basic operations in weaving process. Available: https://web.itu.edu.tr/~berkalpo/Weaving_Lecture/Weaving_Chapter1a_06S.pdf （2025/12/12取得）
Online Fabric Store – Maker’s Mill. (2020). What is Denier? OnlineFabricStore. Available: https://www.onlinefabricstore.com/makersmill/what-is-denier/ （2025/10/11取得）
Patil, R. T., Gulhane, S. S., Raichurkar, P. P., & Basak, S. (2019). Improve productivity of warping by optimization of warping speed and beam pressure. Trends in Textile Engineering & Fashion Technology, 5(2), 635–639.
Pervez, M. N., Yeo, W. S., Lin, L., Xiong, X., Naddeo, V., & Cai, Y. (2023). Optimization and prediction of the cotton fabric dyeing process using Taguchi design-integrated machine learning approach. Scientific Reports, 13, Article 12363.
Ramos-Frutos, J., et al. (2025). Improving the response surface methodology optimization with metaheuristics: A practical approach. Expert Systems with Applications, 297, 129195.
Ribeiro, R., Pilastri, A., Moura, C., Morgado, J., & Cortez, P. (2023). A data-driven intelligent decision support system that combines predictive and prescriptive analytics for the design of new textile fabrics. Neural Computing and Applications, 35, 17375–17395.
RIFA Textile. (2025). AW22 high-speed precise assembly winding machine — Specification. RIFA Textile Machine. Available: https://rifa-textile-machine.en.made-in-china.com/product/hKMEasfrbwcU/China-Aw22-High-Speed-Precise-Assembly-Winding-Machine.html （2025/10/11取用）
Silva, V., & Cardoso, J. S. (Eds.). (2024). Response surface methods—Theory, applications and optimization techniques. BoD–Books on Demand.
Soori, M., Jough, F. K. G., Dastres, R., & Arezoo, B. (2024). AI-based decision support systems in Industry 4.0: A review. Journal of Economy and Technology, 2, 100037.
Sri, D. N., Kottapalli, R., & Almaliki, A. H. (2025). Comparison between response surface methodology and Taguchi method for dyeing process parameters optimization in fabric manufacturing by empirical planning. Scientific Reports, 15(1), 10209.
Stat-Ease. (2026). Central composite design (CCD). Stat-Ease 360 Documentation.
Szpisják-Gulyás, N., Al-Tayawi, A. N., Horváth, Z. H., László, Z., Kertész, S., & Hodúr, C. (2023). Methods for experimental design, central composite design and the Box–Behnken design, to optimise operational parameters: A review. Acta Alimentaria, 52(4), 521–537.
T-Tech Japan Corp. (2021). TTS10S spun sizing machine. Available: https://www.t-techjapan.co.jp/english/download/data/TTS10S_E.pdf （2026/01/21取得）
Textile Learner. (2014). Calculation for direct warping and sectional warping. Textile Learner. Available: https://textilelearner.net/warping-calculation/ （2025/10/11取用）
Textile School. (2012). Weaving calculations. TextileSchool. Available: https://www.textileschool.com/242/weaving-calculations/ （2025/10/11取用）
TT Textiles. (2025). TT Yarn. TT Textiles. Available: https://tttextiles.com/products/yarn/tt-yarn/ （2025/10/11取用）
Ukil. (2025). Apparel beam creel. Available: https://ukil.co/spun-products/spun-sizing/apparel-sizer/apparel-beam-creel/ （2026/03/14取得）
Vetrotex. (2025). Vetrotex. Available: https://www.vetrotextextiles.com/（2026/03/14取得）
Yarn worker. (2018). Geeking out: Doing a loom waste audit. Yarn worker.
Zhang, Y., Hawboldt, K., MacQuarrie, S., Thomas, R., & Gebregiworgis, T. (2024). Alkaline subcritical water extraction of bioactive compounds and antioxidants from beach-cast brown algae (Ascophyllum nodosum). Chemical Engineering Journal, 494, 153109.
Zhou, X., Jiang, T., Zhou, Z., & Hu, X. (2021). Sequential ε-support vector regression based online robust parameter design. Computers & Industrial Engineering, 158, 107391.