以規則為基礎的遞歸選擇性拆卸序列規劃的綠色設計

Abstract

拆卸序列計劃不僅減少產品的生命週期費用，而且還大大地影響環境衝擊。 所以，許多先前的綠色設計研究，集中於廢棄產品的完全拆卸再利用、回收和重新生產有用或貴重的組成成分。要減少環境的影響，許多國家設定某些法規避免進口對環境不友好的產品。考慮綠色設計在拆卸序列計畫階段，參考環境法規是重要的。然而，如果一個特定產品僅回收部分零件或成分，完全拆卸則不實用或不符合經濟效益。選擇性的拆卸序列計劃通常被用於，從產品中拆卸一個或多個成分以重複利用、回收、恢復、和重新製造，以減少對環境的影響。 多數先前的方法列舉所有解答或使用一個隨機方法產生任意解。列舉或隨機方法經常要求巨大計算資源，同時，他們經常可能無法找到符合現實或最佳的方案。這份論文提出一個規則性的遞歸方法以優化在綠色設計中的選擇性拆卸序列。並且，本論文建立某些拆卸規則，其可除去不可行或不切實際的解答。利用這些拆卸規則，能夠減少需要用到的計算資源，並且有效地找出高品質的解。 基於已訂定的規則，在拆除任何零件之前，必須先將限制零件移動的緊固件拆除。然而，在拆除緊固件之前，可能有其它的零件或緊固件也須要移除。在這篇論文裡，設計三個主要的運算功能，對零件和緊固件進行遞歸的移除動作。另外，並不是就每一個單一零件對所有其它零件作幾何限制，本論文只考慮該零件和它的相鄰零件之間的幾何關係。如果取出的零件可以被拆卸，它與相鄰零件之間的幾何關係將被刪除並且更新。最後，此方法可以有效地找出一個在綠色設計中優化選擇性的拆卸序列，並且大大地減少計算時間和空間。

Disassembly sequence planning not only reduces product lifecycle cost, but also greatly influences environmental impact. Therefore, many prior green design research studies have focused on complete disassembly of an end-of-life product to reuse, recycle, recovery, and remanufacturing useful or valuable components. To reduce environmental impact, many countries set up certain regulations to avoid importing environmental unfriendly products. In green design, it is important to consider environmental regulations during the disassembly sequence planning stages. However, complete disassembly is often not practical or cost effective if only a few components will be recovered and recycled from a given product. Selective disassembly sequence planning is usually used to only disassemble one or more components from a product to reuse, recycle, recovery and remanufacturing to reduce environmental impact. Most prior methods either enumerate all solutions or use a stochastic method to generate random solutions. Enumerative or stochastic methods often require tremendous computational resources while, at the same time, they often fail to find realistic or optimal solutions. This thesis presents a rule-based recursive method for finding an optimal heuristic selective disassembly sequence for green design. Based on certain heuristic disassembly rules, the proposed method can eliminate uncommon or unrealistic solutions. Thus, it can greatly reduce computational resources and find high-quality solutions effectively. Based on the defined rules, before any component can be removed, its attached fasteners need to be removed first. However, before the fasteners can be removed, other components or fasteners might need to be removed. In this research, three major functions are developed to handle the recursive removal of components and fasteners. In addition, rather than considering geometric constraints for each pair of components, the developed method only considers geometric relationships between a part and its neighboring parts. If a retrieved part can be disassembled, its geometric relationships with the neighboring parts will dynamically be deleted and updated. As a result, the developed method can effectively find an optimal heuristic selective disassembly sequence while greatly reducing computational time and space.