遗传演算法在船体结构施工程序最适.pdf

遺傳演算法在船體結構施工程序最適化之應用 翁維珠* 洪振發** *國立臺灣海洋大學系統工程暨造船學系 **國立臺灣大學造船及海洋工程研究所 關鍵詞：遺傳演算法、路徑最適化，船殼製造 摘 要 在複雜的船段製造時需要多人同時工作，其配工方式常由領班以等區域簡單劃分每人工作區間再依經驗調整工作負荷， 不易提升整體效率。本文嘗試將施工船段依縱、橫向肋骨劃分成許多施工單元，引用 Okumotor 建議之施工移動法則，建立 施工路徑或施工程序模型，以少數移動決策指標為設計變數，並以遺傳演算法來搜尋最佳工作路徑，使得施工人員閒置比降 至最低，達到最高工作效率之目的。最後，以一包含 6 縱向與 21 橫向大肋骨之船段為研究例，以隨機變數產生 4 種工時分 佈狀況，進行搜尋最佳施工路徑，其結果與理想狀況頗接近。 THE OPTIMIZATION OF WORKING PROCESS FOR SHIP STRUCTURE BY GENETIC ALGORITHMS W.C. Weng** and C.F. Hung** *Department of System Engineering and Naval Architecture National Taiwan Ocean University Keelung, 202 ROC **Department of Naval Architecture and Ocean Engineering National Taiwan University Taipei, 106 ROC Keywords: genetic algorithms, working route optimization, hull construction ABSTRACT In many cases of construction of complex ship structures, it needs many people to work simultaneously. Generally, the works are arranged following the concept of equal working zones, and then adjusted by human experience if the work loading for each worker has remarkable difference. This kind of approach is not easy to achieve a high overall working efficiency for the construction of the ship block. In this paper, we divide the ship block into work cells according to the longitudinal and transverse frames. The moving algorithm to finish the ship block is adopted from