Use dynamic programming to build guillotine cutting mode based on three-dimensional unbounded knapsack

• Main Authors: Cheng-Yen,Wu, 吳政諺
• Other Authors: Shui-Chuan Chen
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/87622678630821828701

碩士 === 國立勤益科技大學 === 工業工程與管理系 === 102 === In the guillotine cutting related study, In recent years, still mainly one-dimensional and two-dimensional problems, The main reason is guillotine cutting belong to NP-hard problem leads to solving problems, which increases the difficulty of solving the factors , allows computers running time is too long, therefore, most of the research in order to facilitate calculation will simplify the high, so a three-dimensional problem into a two-dimensional, even the volume of unit calculation only make the problem into a one-dimensional problem, although this approach efficient, but the practical application difficult, therefore, this study focuses on the study of three-dimensional problem. In this study, the first to First-fit algorithm in the First-Fit Decreasing to sort. Then use the Raster Point to simplify the original data and calculate cutting combination, this method is more accurate than previous data transposed into a series of discrete the way, and can more effectively narrow your search so that a better solution time and accuracy, followed by dynamic programming based on knapsack problem is solved, whereby reached in this study unbounded knapsack unbounded knapsack architecture to solve the problem of the completion of guillotine cutting problem, and to avoid solving the value far from the optimal solution of the problem, and the computation time can have a better performance, then construct systems programming in C #, and validated with 3DLP packages and comparative analysis, proposed in this study to demonstrate the efficiency of the FDR algorithm.