本研究企圖以本文作者發展之modified Artificial Bee Colony（簡稱mABC），針對捷克團隊Martin Vovesný等人發展的玻璃纖維強化塑膠橋面板，進行可靠度為前提的最佳化。

本研究中的玻纖橋面板是由上下兩片長1500 mm寬1060 mm的玻纖板和夾在兩片玻纖板間的玻纖H型樑組成。在不更動元件材料性質和製作程序的情況下，將玻纖板個別厚度及玻纖H型樑數擬為設計變數，外力及材料性質設為隨機變數，盡可能降低製造成本，並控制變形量和Tsai‑Wu failure criterion（蔡吳破壞準則），使玻纖橋面板達成0.99865的可靠度。由於變形量和蔡吳破壞準則均來自耗時的有限元素分析，所以本研究雙管齊下，以叢集式平行運算和mABC分別解決運算速度和最佳化演算法效率的問題。

Gradient‑based optimization（基於梯度的最佳化）或Hessian‑based optimization（基於Hessian的最佳化）一般而言相當有效率，但是比較適合用於局部最佳化，因為它很容易陷入局部最佳解無法自拔。另外，這些方法不適用於不可被微分的函式，當微分的成本過高的時候，可能也會顯得沒效率。population‑based optimization（基於群體的最佳化）雖然有機會避免陷於局部最佳解，不過它通常需要大量的嘗試才能得到足夠好的解。

mABC是一個基於Artificial Bee Colony（蜂群演算法）的最佳化軟體，主要目的在於減少目標式或限制式的計算次數，並取得一個以上具有合理可行精確度的最佳解。與原版蜂群演算法的主要差異有：將連續變數切割成精確度符合需求的離散變數；在新設的階段中使用nearest neighbors（最近鄰居演算法）；在surrogate model（代理模型）上做最佳化，並使用其結果取代原版蜂群演算法scout phase中完全隨機的取樣；調整初始階段中搜尋新解的方式建立更有效的代理模型；能夠依據需要，有限度的放鬆限制式；可以回傳數個相同目標值的最佳解。

本研究以易學易用的直譯式程式語言Python為基礎，盡可能的使用自由軟體、不依賴商業專利軟體。其中，Numpy和Scipy協力提供科學運算不可或缺的各種基礎設施，Scikit‑learn帶來機器學習領域多元化的選擇和可能性，平行運算功能則歸功於IPython。

A temporary bridge panel made of glass fiber reinforced plastic (GFRP) components is subjected to vehicular traffic load. In order to reduce production cost while keeping a specified reliability against variability in material, construction, and load, reliability‑based design optimization (RBDO) is applied to the finite element model of the panel.

Conventional gradient or Hessian based optimization algorithms are thought to be efficient, but they are local optimization algorithms – without a carefully chosen initial point they often fail or deliver sub‑optimal solutions. The calculation of derivatives means a detour is necessary for non‑differentiable problems and these algorithms become less preferable when the derivatives are costly. Current swarm intelligence algorithms are good global optimization propositions, but they are usually slow in convergence.

The goal of this research is to develop an efficient, effective, multi‑purpose optimization algorithm, and use it to determine the best possible GFRP panel designs in computer simulations.

[1] Robinson MJ, Kosmatka JB. Light-Weight Fiber-Reinforced Polymer Composite Deck Panels for Extreme Applications. Journal of Composites for Construction. 2008;12(3):344–354.
[2] Kim H-Y, Park K-T, Jeong J, Lee Y-H, Hwang Y-K, Kim D. A pultruded GFRP deck panel for temporary structures. Composite Structures. 2009;91(1):20–30.
[3] Kim H-Y, Lee S-Y. A steel-reinforced hybrid GFRP deck panel for temporary bridges. Construction and Building Materials. 2012;34:192–200.
[4] Vovesný M, Rotter T. GFRP Bridge Deck Panel. Procedia Engineering. 2012;40:492–497.
[5] Kalnins K, Ozolins O, Jekabsons G. Metamodels in design of GFRP composite stiffened deck structure.
[6] Jin R, Chen W, Simpson TW. Comparative studies of metamodelling techniques under multiple modelling criteria. Structural and Multidisciplinary Optimization. 2001;23(1):1–13.
[7] Karaboga D, Akay B. A comparative study of Artificial Bee Colony algorithm. Applied Mathematics and Computation. 2009;214(1):108–132.
[8] Tsai SW, Wu EM. A General Theory of Strength for Anisotropic Materials. Journal of Composite Materials. 1971;5(1):58–80.
[9] Haldar A, Mahadevan S. Probability, reliability, and statistical methods in engineering design. New York ; Chichester [England]: John Wiley; 2000. 304 p.
[10] Janssen H. Monte-Carlo based uncertainty analysis: Sampling efficiency and sampling convergence. Reliability Engineering & System Safety. 2013;109:123–132.
[11] Viana, F. A. C. Things you wanted to know about the Latin hypercube design and were afraid to ask. Proceedings of the 10th World Congress on Structural and Multidisciplinary Optimization. 2013 May 19.
[12] Naess A, Leira BJ, Batsevych O. System reliability analysis by enhanced Monte Carlo simulation. Structural Safety. 2009;31(5):349–355.
[13] Python Software Foundation. Python 3.4.3. 2015.
[14] Pérez F, Granger BE. IPython: a System for Interactive Scientific Computing. Computing in Science and Engineering. 2007;9(3):21–29.
[15] IPython Project. Using IPython for parallel computing: Overview and getting started. [accessed 2015 Jun 13]. http://ipython.org/ipython-doc/stable/parallel/parallel_intro.html
[16] Van der Walt S, Colbert SC, Varoquaux G. The NumPy Array: A Structure for Efficient Numerical Computation. Computing in Science & Engineering. 2011;13(2):22–30.
[17] Millman KJ, Aivazis M. Python for Scientists and Engineers. Computing in Science & Engineering. 2011;13(2):9–12.
[18] Jones E, Oliphant T, Peterson P, others. SciPy: Open source scientific tools for Python. 2001.
[19] Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel M, Prettenhofer P, Weiss R, Dubourg V, et al. Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research. 2011 Oct.
[20] Thirion B, Duschenay E, Michel V, Varoquaux G, Grisel O, VanderPlas J, granfort alexandre, pedregosa fabian, Mueller A, Louppe G. scikitlearn. 2013.
[21] Octave community. GNU Octave 3.4. 2012.
[22] MATLAB. version 7.10.0 (R2010a). Natick, Massachusetts: The MathWorks Inc.; 2010.
[23] Nocedal J, Wright SJ. Numerical optimization. 2nd ed. New York: Springer; 2006. 664 p. (Springer series in operations research).
[24] Akay B, Karaboga D. A modified Artificial Bee Colony algorithm for real-parameter optimization. Information Sciences. 2012;192:120–142.
[25] Land AH, Doig AG. An Automatic Method of Solving Discrete Programming Problems. Econometrica. 1960;28(3):497.
[26] Welch WJ, Buck RJ, Sacks J, Wynn HP, Mitchell TJ, Morris MD. Screening, Predicting, and Computer Experiments. Technometrics. 1992;34(1):15.
[27] Cornuéjols G. Revival of the Gomory cuts in the 1990’s. Annals of Operations Research. 2007;149(1):63–66.
[28] Tsompanakis Y, Lagaros ND, Papadrakakis M, editors. Structural design optimization considering uncertainties. London ; New York: Taylor & Francis; 2008. 634 p. (Structures and infrastructures series ; v. 1).
[29] De Beer M, Sadzik EM, Fisher C, Coetzee CH. Tyre-Pavement Contact Stress Patterns from the Test Tyres of the Gautrans Heavy Vehicle Simulator (HVS) MK IV+. Proceeding of 24th Annual Southern African Transport Conference and Exhibition. 2005 July 11-13.
[30] Majumdar PK, Lesko JJ, Cousins TE, Liu Z. Conformable Tire Patch Loading for FRP Composite Bridge Deck. Journal of Composites for Construction. 2009;13(6):575–581.
[31] Zhou A, Coleman JT, Temeles AB, Lesko JJ, Cousins TE. Laboratory and Field Performance of Cellular Fiber-Reinforced Polymer Composite Bridge Deck Systems. Journal of Composites for Construction. 2005;9(5):458–467.