改进非支配排序进化算法在下料问题中的应用  

Application for cutting stock problems based on improved non-dominated sorting evolution algorithm

作　　者：阚方 刘林[1,2] 方昶[1] 裴军[1] 郑明月[1] 

机构地区：[1]合肥工业大学管理学院,合肥230009 [2]合肥工业大学管理学院过程优化与智能决策教育部重点实验室,合肥230009

出　　处：《计算机工程与应用》2014年第15期244-248,共5页Computer Engineering and Applications

基　　金：国家自然科学基金重点基金(No.71231004);国家自然科学基金(No.71171071);安徽省高校省级自然科学研究项目(重点)(No.KJ2011A215)

摘　　要：针对一维下料问题,提出了减少废料、减少下料设置时间和减少可回收余料的三目标优化模型,用改进的非支配排序进化算法求出问题的Pareto最优解集,运用逼近理想解方法从解集中选出一个满意解作为下料方案,各优化目标的权重用CRITIC法算出。仿真实验证明了所提出的方法可以有效解决该类多目标下料问题。A multi-objective cutting stock problem is studied. The objects are to minimize the non-reuse materials, the cutting waiting time and usable leftovers. A decision-making method for this cutting stock problem is designed. The Pareto-optimal set is gotten by using the improved non-dominated sorting heuristic evolutionary algorithms. The weight of every object is calculated by the CRITIC method. A satisfactory solution, which is regarded as cutting scheme, is found from the Pareto-optimal set by employing the multi-attribute decision making method. The results of experimentation show that the proposed method is effective.

关 键 词：一维下料 多目标优化 进化算法 多属性决策 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]