基于改进ICA算法的LBFFSP问题研究  

Research on Flexible Flow Shop Scheduling Problem with Limited Buffer Based on an Improved ICA Algorithm

作　　者：韩忠华[1,2] 孙越[1] 史海波[2] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168 [2]中国科学院沈阳自动化研究所,辽宁沈阳110016

出　　处：《信息与控制》2017年第4期474-482,共9页Information and Control

基　　金：国家自然科学基金资助项目(61503259);辽宁省科技厅项目(201602608)

摘　　要：为了解决带有限缓冲区的柔性流水车间排产优化问题(Limited-Buffer Flexible Flow-shop Scheduling Problem,LBFFSP),首先建立LBFFSP的数学模型,提出了一种改进帝国竞争算法(improved imperialist competitive algorithm,IICA)作为全局优化算法,在标准帝国竞争算法基础上,引入模拟退火思想,扩大算法搜索范围,并加入离散化处理操作、改革操作、以及精英个体保留策略三处改进.为进一步提高算法搜索最优解效率,设计了一种基于优化目标的初始种群建立方法,并加入基于汉明距离的个体选择机制,以提高初始种群中初始解的质量.设计仿真实验,对算法中的参数进行分析探讨,确定最佳参数值.最后通过实例测试,将IICA算法与其他算法进行对比研究,验证了IICA算法对于解决柔性流水车间有限缓冲区的排产优化问题的有效性.To solve the limited-buffer flexible-flow shop scheduling problem( LBFFSP),a mathematical model of the LBFFSP is established,and an improved imperialist competitive algorithm( IICA) is proposed as the global optimizing algorithm. The idea of simulated annealing is introduced to expand the scope of the search algorithm over the standard imperialist competitive algorithm,which joins three modifications,namely,the discretization processing operation,the reform operation,and the elite individual retention strategy. To further improve the efficiency of the algorithm in searching for the optimal solution,the initial population establishment method based on optimization objective is designed. The individual selection mechanism is added to improve the initial solution quality of the initial population by using Hamming distance. The algorithm parameters are analyzed to determine the optimum parameter values through simulation experiments. The effectiveness of the IICA in solving the limited-buffer flexible flow shop scheduling problems is determined by comparing the example test results of IICA with those of other algorithms.

关 键 词：有限缓冲区 改进帝国竞争算法 构建初始种群 汉明距离 

分 类 号：TH165[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]