以廣義基因演算法搜尋黏滯阻尼器與鐵磁性薄膜沉積之最佳化參數

Abstract

最佳化分析是所有工程分析上所欲求之最終狀態，不過最佳化過程常因領域的不同而存有不同的難度。本文將嘗試改良基因演算法使之為一適合跨領域研究最佳化之數值方法，並分別以材料及結構領域進行研究。材料方面，以鐵磁性薄膜為對象，其中入射角度、基板溫度及入射能量為分析參數，並運用分子動力學模擬濺鍍過程，以進行薄膜表面微觀結構分析；在結構被動控制方面，剪力屋架中所安裝之黏滯性液流阻尼器為主要分析目標，分別來探討廣義基因演算法之可行性。

Optimum design is the final searching state on each engineering, but the optimum handle process has different analytic difficult. This studying attempts to improve the genetic algorithm that will turn into the numerical method of optimum analysis. This studying will verify the method in material and structure domain. In material domain, the ferromagnetism thin film is analytic project. Molecular dynamics (MD) simulation is employed to investigate the surface topography of thin film produced by the sputtering process for different parameters such as substrate temperature, incident energy and incident angle. In structure domain, the viscous-damper is analytic project. This studying will search the optimum parameters of viscous-damper and thin film with generalized genetic algorithm.