JSWENO 算則為 Jiang and Shu 於 1996 年所發表的 WENO5 算則，
組合了三階精度的 ENO 算則，以定義之平滑指示器使在通量的計算 上，平滑區能盡量保持五階的精度，高震波區則令含有不連續解之計 算元的權重盡量減小，以維持數值傳遞不震盪的原則；ZSWENO 算 為 Zhang and Shu 於 2007 年針對 JSWENO 算則於強震波區域的改進 方法，捨棄 JSWENO 算則定義之平滑指示器中含二次微分項的部份，
作為修正平滑指示器以增強在強震波區域的收斂性；而將 ZSWENO 算則應用於一維淺水波方程式中，發現在某些案例中會有較不好的模 擬結果，似乎有過度收斂的情形發生。
因此本研究針對 ZSWENO 算則，加入修正係數 以重新定義平 滑指示器，並藉由類神經網路假設其構成平滑指示器之五點與修正係 數 間的關係函式，以 ZSWENO 模擬較差的案例作為學習案例(穩態 流流經障礙物)、與案例解析解的誤差作為學習好壞的依據，結合遺 傳演算法的優選特性來學習，最終得到一 AWENO 的算則。
6-1 結論
經 過 各 種 一 維 淺 水 波 方 程 式 的 案 例 詴 驗 ， AWENO 算 則 與 JSWENO 算則、ZSWENO 算則之模擬結果比較分析後，可得以下之 結論：
(1) AWENO 算則之類神經網路只有對穩態流流經障礙物一案例做學 習，而此次學習結果，著重於反應其不連續面上的不震盪性，因 此經學習之類神經網路，在簡單平滑曲線微分案例的通量計算上，
其精度的表現為三種算則中之最低。或許在學習過程中，可加入 如簡單平滑曲線微分案例等對通量計算的精度要求較為重要的學 習案例，以提昇 AWENO 算則在精度的表現。
(2) 經無劇烈通量變化之一維淺水波方程詴驗案例的測詴，三種算則 所模擬結果與誤差都大致相同，似乎精度的保持對於無劇烈通量 變化一維淺水波方程的數值通量計算上，影響性較不這麼重要。
(3) 在水面分離與潰壩案例的模擬表現上，AWENO 算則大致介於 JSWENO 與 ZSWENO 算則之間，對於乾濕點交界處的收斂性較 JSWENO 算則來得好，較 ZSWENO 算則來得差；而在對於水位線 變化的連接處理上，反之較 ZSWENO 算則來得好，較 JSWENO 算則來得差。
(4) AWENO 算則在乾濕點交界處的收斂處理上，水面線會有向右方 些許偏傾的現象，而就模式的穩定性而言，最好能保持其計算上 的對稱性。也因此這不對稱性的收斂，運用於更複雜與更長久時 間的案例運算上，或許會出現數值通量傳遞上的誤差，因而產生 數值的震盪。
(5) 對於一般的明渠案例模擬而言，案例通常是由平滑區、緩變段與 震波區組合而成，極少有單一高強度震波的情況，而由模擬的結 果可知 JSWENO 算則的應用廣度與穩定性相對來地高，以一般淺 水波方程式案例而言，選用 JSWENO 算則模擬通常都可得到不錯 之模擬結果。
6-2 建議
1. 在學習過程中，除了單純處理類神經網路的運算外，還需合併水 理模式的運算，此外一學習循環又包括了 n 組的基因組合，這使 得在學習上需耗費大量的時間。若能透過平行運算同時間處理 n 組基因組的計算，應能大幅節省學習所花費的時間使之學習更有 效率。
2. 利用類神經網路來擬合所假設未知的關係函式，就理想而言透過 無數次案例、無數次反覆的學習，此擬合的關係函式應能漸趨於 一常態穩定的型態，自訂類神經網路的應用廣度與穩定性也能更 佳；但因類神經網路的不確定性，輸入與輸出中間傳遞的過程難 以用物理原則加以解釋，加上要考慮設計網路之架構與規模是否 能反應其未知函式的變化、學習需耗費大量時間等缺點，因此在 現實面的實行上有一定的難度。故若要繼續探討加權基本不震盪 法收斂性之改良，由於類神經網路是假設其構成平滑指示器之五
點與修正係數 間的關係函式，在增加學習案例的同時，或許 可以利用類神經所學習之結果，結合物理概念，嘗詴建構出一套 能以物理解釋之修正算則。
參考文獻
1. Abgrall R. (1994), “On essentially non-oscillatory schemes on unstructured meshes: analysis and implemen-tation.” J. Comput.
Phys., Vol. 114, pp. 45-58.
2. Berm dez A., Dervieux A., D sid ri J.A., and V zquez M.E. (1998),
“Upwind schemes for the two-dimensional shallow water equations with variable depth using unstructured meshes.” Comput.Methods Appl.Mech.Eng.,155, 49.
3. Berm dez A. and V´azquez M.E. (1994), “Upwind methods for hyperbolic conservation laws with source terms.” Computers and Fluids., Vol. 23, pp. 1049-1071.
4. Botchorishvilli R., Perthame B., and Vasseur A. (2000), “Equilibrium schemes for scalar conservation laws with stiff sources.”
Mathematics of Computation, Vol. 72, pp. 131-157.
5. Bouchut F. (2004), “Nonlinear stability of finite volume methods for hyperbolic conservation laws and well-Balanced schemes for sources.” Birkhauser Verlag, Germany.
6. Chaves P. and Chang F.-J. (2008), “Intelligent reservoir operation system based on evolving artificial neural networks.” Advances in Water Resources, Vol. 31, pp. 926-936.
7. Chen K.-H. and Pletcher R.H. (1991), “Primitive variable, strongly implicit calculation procedure for viscous flows at all speeds.” AIAA J., Vol. 29(8), AUGUST., pp. 1241-1249.
8. Chinnayya A. and LeRoux A.Y.(1999), “A new general Riemann solver for the shallow-water equations with friction andtopography.”
Available in the conservation law preprint server http://www.math.ntnu.no/conservation/.
9. Choi, H. and Liu J.-G. (1998), “The reconstruction of upwind fluxes for conservation laws: its behavior in dynamic and steady state calculations. ” J. Comput. Fluids., Vol. 144, pp. 237-256.
10. Crnjaric-Zic N., Vukovic S., and Sopta L. (2004), “Extension of ENO and WENO schemes to one-dimensional sediment transport equations.” J. Comput. Fluids., Vol. 33, pp. 31-56.
11. Daliakopoulos I.N., Coulibaly P., and Tsanis, I.K. (2005).
“Groundwater level forecasting using artificial neural networks.”
Journal of Hydrology, Vol. 309(1-4), pp. 229-240.
12. Despres B. and Lagoutiere F. (2001), “Contact discontinuity capturing schemes for linear advection, compressible gas dynamic.”
Journal of Scientific Computing, Vol. 16, pp. 479-524.
13. Einfeldt B. (1988), “On Godunov-type methods for gas dynamics.”
SIAM J. Numer. Anal., Vol. 25, pp. 294-318.
14. Fatemi E., Jerome J., and Osher S. (1991), “Solution of the hydrodynamic device model using high order non-oscillatory shock capturing algorithms.” IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol. 10, pp. 232–244.
15. French M.N., Krajewski W.F., and Cuykendall R.R. (1992),
“Rainfallforecasting in space and time using a neural network.”
Journal of Hydrology, Vol. 137, pp. 1-31.
16. Friedrich O. (1998), “Weighted essentially non-oscillatory schemes for the interpolation of mean values on unstructured grids.” J.
Comput. Phys., Vol. 144, pp. 194-212.
17. Gerritsen, M. and Olsson, P. (1998), “Designing an efficient solution strategy for fluid flows.” J. Comput. Phys., Vol. 147, pp. 293-317.
18. George D.L. (2004), “Numerical approximation of the nonlinear shallow water equation : a godunov-type scheme, master’s thesis.”
University of Washington, Seattle, WA.
19. Goldberg D.E. (1989),“Genetic algorithms for search, optimization, and machine learning.” Addison-Wesley Longman Publishing Co., Inc.
20. Gosse L. (2001), “A well-balanced scheme using non-conservative products designed for hyperbolic systems of conservation laws with source terms.” Math. Mod. Meth. Appl. Sci., Vol. 11, pp. 339-356.
21. Goutal N. and Maurel F. (1997), “Proceedings of the 2nd workshop on dam-breakwave simulation,technical report he-43/97/016/a.”
de France, Laboratoire National d’Hydraulique, Groupe Hydraulique Fluviale.
22. Greenberg J.M. and LeRoux A.Y. (1996), “A well-balanced scheme for the numerical processing of source terms in hyperbolic equations.”
SIAM Journal on Numerical Analysis, Vol. 33, pp. 1-16.
23. Harten A. (1983), “High resolution schemes for hyperbolic conservation laws.” J. Comput. Phys., Vol. 49, pp. 357-393.
24. Harten A. and Osher S. (1987), “Uniformly high-order accurate non-oscillatory schemes.” SIAM J. Numer. Anal., Vol. 24, pp.
279-309.
25. Harten A. (1989), “ENO schemes with subcell resolution.” J. Comput.
Phys., Vol. 83, pp. 148-184.
26. Henrick A.K., Aslam T.D., and Powers J.M. (2005), “Mapped weighted essentially non-oscillatory schemes.” J. Comput. Phys., Vol.
207, pp. 542-567.
27. Holland J. (1975), “Adaptation in natural and artificial systems.”
The University of Michigan Press, Ann.
28. Hsieh T.J., Wang C.-H., and Yang J.-Y. (2008), “Numerical experiments with several variant WENO schemes for the euler
Vol. 58, pp. 1017-1039.
29. Hubbard M.E. and Garcia-Navarro P. (2000), “Flux difference splitting and the balancing of source terms and flux gradients.” J.
Comput. Phys., Vol. 165, pp. 89-125.
30. Jain, S.K., Das A., and Srivastava, D.K. (1999), “Application of ANN for reservoir inflow prediction and operation.” J. Water Resour. Plan.
Manage., Vol. 125(5), pp. 263-271.
31. Jenny P.and Muller B. (1998), “Rankine–Hugoniot–Riemann solver considering source terms and multidimensional effects.” J. Comput.
Phys., Vol. 145, pp. 575-610.
32. Jiang C.-S. and Shu C.-W. (1996), “Efficient implementation of weighted ENO schemes.” J. Comput. Phys., Vol. 126, pp. 202-228.
33. Jiang G.-S. and Tadmor E. (1998), “Nonoscillatory central schemes for multidimensional hyperbolic conservation laws.” SIAM Journal on Scientific Computing, Vol. 19, pp. 1892-1917.
34. Jin S. (2001), “A steady-state capturing method for hyperbolic systems with geometrical source terms.” Mathematical Modelling and Numerical Analysis (M2AN), Vol. 35, pp. 631-645.
35. Kuo A.-C. and Polvani L.M. (1999), “Wave vortex interaction in rotating shallow-water- part I.” Journal of Fluid Mechenics, Vol. 394, pp. 1-27.
36. LeVeque R.J. (1990), “Numerical methods for conservation laws.”
Birkhauser Verlag, Basel.
37. LeVeque R.J. (1998), “Balancing source terms and flux gradients on high resolution Godunov methods: the quasi-steady wave propagation algorithm.”, J. Comput. Phys., Vol. 146, pp. 346–365.
38. LeVeque R.J. (2002), “Finite volume methods for hyperbolic problems.” Cambridge University Press, UK.
39. Liska R. and Wendroff B. (1999), “2-Dimensional shallow-water
equations by composite schemes.” International Journal for Numerical Method in Fluid, Vol. 30, pp. 441-479.
40. Liu X.D., Osher S., and Chan T. (1994), “Weighted essentially nonoscillatory schemes.” J. Comput. Phys., Vol. 115, pp. 200-212.
41. Liu X.D. and Osher S. (1998), “Conex ENO high order multi-dimensional schemes without field by fiele decomposition orstaggered grids,” J. Comput. Phys., Vol. 142, pp. 304-330.
42. Loke E., Warnaars E.A., Jacobson P., Nelen F., and DoC'eu M.A.
(1997), “Artificial neural networks as a tool in urban storm drainage.” Wat. Sci. Tech., Vol. 36, pp. 101-109.
43. Lorrai M. and Sechi G.M. (1995). “Neural nets for modeling rainfall-run transformations. ” Wat. Resour. Man. Vol.9, pp.299-313.
44. McCulloch W.S. and Pitts W. (1943),“A logical calculus of the ideas immanent in nervous activity.” Bulletin of Mathematical Biophysics, Vol. 5, pp. 115–133.
45. Minion M.L. and Brown D.L. (1997), “Performance of under-resolved two-dimensional incompressible flow simulations II.”
J. Comput. Phys., Vol. 138, pp. 734-765.
46. Ollivier-Gooch C.F. (1997), “Quasi-ENO schemes for unstructured meshes based on unlimited data-dependent least-squires reconstruction,” J. Comput. Phys., Vol. 133, pp. 6-17.
47. Roe P. L. (1986), “Characteristics-based upwind scheme for the euler equations.” Annual Review of Fluid Mechanics, Vol. 18, pp.
337-365.
48. Rumelhart D.E., Hinton G.E., and Williams R.J. (1986), “Learning internal representations by error propagation.” in D.E. Rumelhart and J.L. McClelland eds. Parallel Distributed Processing:
Explorations in the Microstructure of Cognition, Vol. 1, pp. 318-362.
rainfall-runoff model using an artifical neural network.” Journal of Hydrology, Vol. 216(1), pp. 32-55.
50. Shigidi A. and Garcia L.A. (2003), “Parameter estimation in groundwater hydrology using artificial neural networks.” Journal of Computing in Civil Engineering, Vol. 17(4), pp. 281-289.
51. Shu C.-W. (1990), “Numerical experiments on the accuracy of ENO and modified ENO schemes.” Journal of Scientific Computing, Vol. 5, pp. 127-149.
52. Shu C.-W. (1997), “Essentially non-oscillatory and weighted essentially non-oscillatory schemes for hyperbolic conservation laws.”
Institute for Computer Applications in Science and Engineering NASA Langley Research Center Hampton, VA.
53. Shu C.-W. (1998), “Essentially non-oscillatory and weighted essentially non-oscillatory schemes for hyperbolic conservation laws.”
Advanced Numerical Approximation of Nonlinear Hyperbolic Equations, Lecture Notes in Mathematics, Vol. 160, Springer, Berlin/New York.
54. Shu C.-W. and Osher S. (1988), “Efficient implementation of nonoscillatory shock capturing scheme.” J. Comput. Phys., Vol. 77, pp. 439-471.
55. Shu C.-W. and Osher S. (1989), “Efficient Implementation of Nonoscillatory Shock Capturing Schemes II.” J. Comput. Phys., Vol.83, pp.32-78.
56. Suresh, A. and Huynh (1997), “Accurate monotonicity-preserving schemes with Runge-Kutta time stepping.” J. Comput. Phys., Vol.
136, pp. 83-99.
57. V´azquez-Cendon M.E. (1999), “Improved treatment of source terms in upwind schemes for the shallow water equations in channels with irregular geometry.” J. Comput. Phys., Vol. 148, pp. 497-526.
58. Van Leer B. (1974), “Towards the ultimate conservative difference scheme Ⅱ, monotonicity and conservation combined in a second order scheme.” J. Comput. Phys., Vol. 14, pp. 361-470.
59. Van Leer B. (1979), “Towards the ultimate conservative difference scheme Ⅴ, a second order sequel to Godunov’s method.” J. Comput.
Phys., Vol. 32, pp. 101-136.
60. Vukovic S. and Sopta L. (2002), ”ENO and WENO schemes with the exact conservation property for one-dimensional shallow water equations.” J. Comput. Phys., Vol. 179, pp. 593-621.
61. Werbos P. (1974), “Beyond regression: new tools for prediction and analysis in the behavioral sciences.” PHD thesis, Harvard, Cambridge, MA.
62. Xing Y. and Shu C.-W. (2005), “High order finite difference WENO schemes with the exact conservation property for the shallow water equations.” J. Comput. Phys., Vol. 208, pp. 206-227.
63. Xu Z. and Shu C.-W. (2005), “Anti-diffusive flux corrections for high order finite difference WENO schemes.” J. Comput. Phys., Vol. 205, pp. 458-485.
64. Yang J.-Y. and Hsu. C.-A. (1992), “High-resolution, nonoscillatory schemes for unsteady compressible flows.” AIAA Journal, Vol. 30(6), pp. 1570-1575.
65. Yang J.-Y., Yang S.-C., Chen Y.-N., and Hsu C.-A. (1998), “Weight ENO schemes for the three-dimensional incompressible Navier-Stoke equations.” J. Comput. Phys., Vol. 146, pp. 464-487.
66. Yee H.-C. (1989), “A class of high-resolution explicit and implicit shock capturing methods.” NASA TM-101088.
67. Zhang B. and Rao S.-G. (2000), “Prediction of watershed runoff using Bayesian concepts and modular neural networks.” Water
Resources Research, Vol. 36(3), pp. 753-762.
68. Zhang S. and Shu C.-W. (2007), “A new smoothness indicator for the WENO schemes and its effect on the convergence to steady state solutions.” Journal of Scientific Computing, Vol. 31, Nos. 1/2.
69. Zhou J.-G., Causon D.M., Mingham C.G., and Ingram D.M. (2001),
“The surface gradient method for the treatment of source terms in the shallow-water equations.” J. Comput. Phys., Vol. 168, pp. 1-25.
70. 王進德、蕭大全(1994)，「類神經網路與模糊控制入門」，全華圖
經網路」，國立台灣大學土木工程學研究所碩士論文。
79. 黃國源(2003)，「類神經網路與圖型識別 第二版」，維科圖書公 司。
80. 張東炯(2000)，「類神經網路於土路流發生預測模式之研究」，台 灣水利，第四十八卷，第二期，第 92∼98 頁。
81. 曾國源(2003)，「以類神經網路架構土石流預警系統之研究」，國 立臺灣大學生物環境系統工程學系暨研究所博士論文。
82. 葉怡成(2001)，「應用類神經網路 第三版」，儒林圖書公司。
83. 葉怡成(2001)，「類神經網路模式應用與實作 第七版」，儒林圖書 公司。
84. 謝蒼仁(2009)，「隱式逆擴散加權基本不震盪算則的發展及其應 用」，國立台灣大學工學院機械工程學研究所博士論文。
85. 羅華強(2001)，「類神經網路–MATLAB 的應用」，清蔚科技公司。
附錄 A. 類神經網路
類神經網路一種基於腦與神經系統研究所啟發的資訊處理技術，
使用大量的相連人工神經元來模仿生物神經網路的能力，其運算單元 由類似生物腦神經元的簡單處理器組成，而處理器間的連結則是模仿 神經元間的突觸，反映了生物腦神經系統的基本特性。
A-1 生物神經元模型
生物神經網路是由巨量的神經細胞(neuron)，或稱神經元所組成，
其構造包括(圖 A- 1)：
(1) 神經核(soma)：神經細胞呈核狀的處理機構。
(2) 軸索(神經軸)(axon)：神經細胞呈軸索狀的輸送機構。
(3) 樹突(神經樹)(dendrites)：神經細胞呈樹枝狀的輸出入機構。
(4) 突觸(神經節)(synapse)：神經樹上呈點狀的連結機構。
(4) 突觸(神經節)(synapse)：神經樹上呈點狀的連結機構。