- 學位論文
利用啟發式演算法設計不規則形狀物料的切割系統
Applying heuristics algorithm to design a cutting stock system for arbitrary shape objects
摘要
製造過程中的物料舉凡玻璃、木板、鋼板、鋁片…等物料，都必需經過切割才能再進行後續的製程和組裝程序。切割技術中最重要的一環即是如何規劃的切割問題。近年來全球經濟衰退、通貨膨脹而原物料價格不斷上漲，造成企業的生產成本增加。必須減少物料的使用浪費來減輕生產成本的壓力。傳統切割問題只能解決方型物件切割，但實際應用中需要切割出的物件大多都是不規則的形狀並且切割的物料可能有破損。因此，本研究著重於開發一套可以使用有缺陷物料進行不規則物件的切割系統。透過切割近似實際物件形狀的不規則物件來減少廢料的產生，直接切割有缺陷物料增加使用效率，並進行切割之後還可以重複切割，將剩餘的物料再進一步的利用，讓物料的浪費達到最少。系統以數學形態學進行切割位置規劃，並利用基因演算法尋找最佳的物件切割順序。希望透過本研究開發的系統，將物料的浪費達到最少。
並列摘要
Any raw material from the manufactory industry has to be proper tailored before it comes to any post processing procedures (such as mass production, assembling and delivery). The tailoring processes, generally referred as cutting stock, are very common industrial applications worldwide. The most important for cutting stock application is to provide the cutting patterns for manufactory. Due the global recession and the inflation, the price of the raw material raises dramatically. This has caused the cost problem to many businesses. How to reduce the waste and provide automatic cutting patterns is particularly important. The traditional cutting stock can only handle the rectangular shapes; however, the true shape objects are the most common in real world application. Also, the traditional cutting stock can not handle the piece with defect, and can not allow the re-cutting procedure. To overcome these problems, this study is aimed to develop a cutting stock for irregular shape cutting which can handle the defect and re-nesting procedure. It is the goal of this study to reduce the raw material waste and provide the direct cutting patterns to enhance the automation. This study utilizes the morphology theory as the basis and the genetic algorithm as a tool for optimization search to yield the best cutting patterns.