A Study on Solving 1D-RCSP of Periodic-time Orders Using Genetic Algorithm

• Main Authors: Shih-ChinLiao, 廖世欽
• Other Authors: Lih-Chyun Shu
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/59943262707551882624

碩士 === 國立成功大學 === 會計學系碩博士班 === 101 === The main idea of this thesis is related to Residual Cutting Stock Problem (RCSP), and we focus on minimizing trim loss and cutting costs in factory subsequent process. Since the demands for orders are periodical, that is, the distributions of orders are similar in seasons among different years, our study explores how to adjust the length and amount of residual stock which is conducive to trim loss and minimize cost in the entire continuous cutting process by using last year’s orders to predict subsequent year’s orders. When arranging cutting plans, we assume that the length and orders of residual stocks are as much as possible the same and these residual stocks will never produce trim loss in subsequent cutting process, thus reducing the whole cutting costs. In problem solving, we use genetic algorithm to solve the problem proposed by this thesis.