頁籤選單縮合
題名 | 基因演算法應用於營建業計數型抽樣驗收計畫之研究=Applying Genetic Algorithm on Construction Quality Attribute Acceptance Sampling Plan |
---|---|
作者姓名(中文) | 鄭道明; 馮重偉; 郭陽春; | 書刊名 | 中國土木水利工程學刊 |
卷期 | 15:3 2003.09[民92.09] |
頁次 | 頁615-621 |
分類號 | 441.3 |
關鍵詞 | 基因演算法; 計數型驗收; 作業特性曲線; Genetic algorithm; Attribute acceptance; OC-Curve; |
語文 | 中文(Chinese) |
中文摘要 | 品質抽樣驗收計畫對於選擇具有品質保證的建材於營建工程中施作扮演重要角色,其中「計數型抽樣驗收」具有應用簡易的特性,適用於營建工程中欲避免繁複統計計算的場合,可以快速協助現場工程人員決定營建工程產品應該被「接受」或「拒收」。然而,由於計數型抽樣驗收的求解參數式為離散數學模式,統計學公式無法直接求解,雖然已有研究採用窮盡搜索的方式求解,能確定解答的品質為最佳,但是對於更寬廣的求解範圍,則需要考慮更具效率的演算方式。本研究採基因演算法利用業主和承包商於合約內所擬定之品質要求門檻 (AQL和RQL) 的高低與欲承擔之風險 ( Risk和 Risk) 代入研究模式,迅速求解抽樣驗收的參數值,進而達成經濟之參數設計,並建立具親和力的求解操作程式,使得抽樣驗收計畫變得具體可行。 |
英文摘要 | An acceptance sampling plan plays an important role in quality assurance in construction, and applies statistics to specify the requirements of (1) how many measurements are needed, and (2) how to make an acceptance or rejection decision based on measured data. A good acceptance sampling plan can not only reduce the cost of construction inspection, but also increase the accuracy of acceptance decision. This research proposes a quality attribute acceptance sampling system that applies Genetic Algorithms (GA) for facilitating the selection of sample size and acceptance criteria. Case runs for verifying the effects of adopting GA to enhance optimization capability of construction quality attribute acceptance sampling show a positive conclusion. In addition, a user friendly computer program using Visual Basic is presented in this paper. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。