透過您的圖書館登入
IP:18.222.35.77
  • 學位論文

旱季時感潮河川自淨的最佳化操作方法研究

The study on the method of optimal operation of self-purification for a tidal river in dry seasons

指導教授 : 黃良雄
共同指導教授 : 黃仲偉

摘要


考慮旱季時感潮河川由於潮汐入流量與潮汐出流量相當,造成河川自淨能力相較於豐水期時為差。若輔以閘門的操作不失為有效改善水質的方法之一,故為研究旱季時感潮河川自淨的最佳化操作方法,以力學的觀點初步建立一個理想化的具有使用閘門開關來操作河川自淨能力之參考模型,並完成其半解析解。而為了後續最佳化研究計算效率之考量,藉由合理的時間尺度關係,建立模型近似解的求解方法,其具有能相對快速求解,且在數學上是相對穩定的特性。由半解析解與近似解之比對,顯示此近似解能有效控制誤差。 本研究共提出了三種旱季時感潮河川自淨的最佳化操作方法,分別為靜態閘門控制最佳化、動態閘門控制最佳化(不考慮操作時間之間的交互作用)以及動態閘門控制最佳化(考慮操作時間之間的交互作用)。其中,最佳化方法的部分,以本研究設計的反應曲面法應用,使得其相較於以往方法而言,較能處理設計變數與目標函數間的關係非單調變化,但其仍存有特定趨勢之問題,使最佳化求解之結果仍能盡量逼近全域真實最佳解。而案例應用結果顯示,三種最佳化操作方法,即不同策略的最佳化操作,操作上限制較少的情況,確實能有較好的最佳化結果。 最後,應用本研究發展的模型和最佳化方法,以汙染物位置或汙染物濃度最大值作為目標函數,分別求得在不同的最佳化策略前提下,如何得到最佳的感潮河川自淨能力。由結果比對可發現,反應曲面直接掌握系統整體之結果,包含移流延散機制、濃度分佈梯度、邊界條件以及不同時間操作交互作用等效應,將錯綜複雜之過程以一簡單函數關係取代,以有效求得不同的最佳化策略下,感潮河川自淨能力最佳化的操作方法。

並列摘要


Considering that the tidal rivers during the dry seasons only have the small flow in the river, the self-purification effect is worse than that in high flow period. To study the optimal operation for this situations, based on mechanics, the idealized model with operational weirs is established with a semi-analytical solution. And, a rational time scale is employed to construct an efficiency and stable approximate solution. By the comparison between the semi-analytical solution and the approximate solution, the consistency is confirmed. Next, three different optimal operation methods are designed, named as static weir operation optimization, dynamic weir operation optimization (without different time operations interaction) and dynamic weir operation optimization (with different time operations interaction). For finding the optimal operation, the designed method with using response surface methodology is applied, which is suitable for problems with some varieties between the design variables, but still has some specified trends for the objective function, to find the approximate optimal operation, which could be near the real optimal one, for the self-purification of the river. Finally, the developed models and optimization methods are applied to the problems with taking the contaminant position or maximum concentration as the objective function, and the three different optimal operation methods are used to find the different optimal results of the selected cases. The results show that the response surfaces have included the mixed effect combined by the advection-dispersion effect, the gradient of concentration distribution, the diffusion in the closed region, boundary conditions, and the interactions between different time operations, and the optimal operation for the self-purification of the river to each case could be found.

參考文獻


1. Alvarez-Vázquez, L. J., Martínez, A., Vázquez-Méndez, M. E., & Vilar, M. A. (2009). An application of optimal control theory to river pollution remediation. Applied Numerical Mathematics, 59(5), 845-858.
2. Ames, W. F. (Ed.). (1965). Nonlinear partial differential equations in engineering (Vol. 18). Academic press.
3. Aris, R. (1956). On the dispersion of a solute in a fluid flowing through a tube. Proc. R. Soc. Lond. A, 235(1200), 67-77.
4. Arora, J. S. (2004). Introduction to optimum design. Elsevier.
5. Arora, J. S. (2010). Formulating Design Problems as Optimization Problems. Encyclopedia of Aerospace Engineering.

延伸閱讀