本論文應用調和搜尋演算法於結構最佳化設計中。調和搜尋演算法是模仿樂團演奏練習將曲目調整至最協調的現象引用到最佳化演算系統當中，而發展出一套啟發式演算法。樂團演奏練習中每一個樂器代表著一個設計變數，樂團藉由一次又一次的練習達到最完美的表演，就如同演算法藉由一次又一次的迭代使得目標函數達到最佳值。調和搜尋演算法利用調和記憶考慮機率和調音機率決定搜尋方式，因此調和搜尋演算法可以作隨機全域搜尋，可避免落入區域最佳解。本研究將ANSYS有限元素分析軟體中的APDL語法與FORTRAN程式結合成一系統程式，並以六種不同的範例執行結構最佳化設計。範例中將結構最佳化問題轉為數學函數，以結構輕量化為目的，再利用調和搜尋演算法對結構系統執行最佳化設計。由數值分析範例之結果，顯示調和搜尋演算法在結構最佳化設計上可得到不錯的結果。 The Harmony Search (HS) Algorithm was applied to the optimum design of structures in this study. The Harmony Search algorithm was conceptualized using the musical process of searching for a perfect state of harmony. The best natural musical performance occurs when a musician searches a best state of harmony, just as the optimization process searches to find a global solution. The set of sounds are produced by a group of instruments, just as objective function is determined by the set of the values produced by design variables . The sound for better harmony can be improved through practice over and over again, just as the values for better objective function can be improved through iteration by iteration. Therefore the heuristic algorithm derived from an artificial phenomenon found in musical performance, in essence the process of searching the better harmony, can be implemented. The Harmony Search algorithm does not require the setting of initial values of decision variables and uses the random search based on the Harmony Memory Considering Rate (HMCR) and Pitch Adjusting Rate (PAR) for guiding a global search. These features increase the flexibility of the HS algorithm and obtain better solutions efficiently.The FORTRAN and APDL of ANSYS software are integrated into a systematic Harmony Search optimization program. The optimization problem can be transformed into a mathematical function. Minimum weight design will be developed in six numerical examples. Then the optimum deign of structures can be obtained by Harmony Search algorithm. The results of Harmony Search algorithm are better than other references in the examples.
