计算机科学 ›› 2013, Vol. 40 ›› Issue (Z6): 67-69.

• 智能算法与优化 • 上一篇    下一篇

面向装箱问题的量子遗传优化算法

郭晶,陈贤富   

  1. 中国科学技术大学电子科学与技术系 合肥230027;中国科学技术大学电子科学与技术系 合肥230027
  • 出版日期:2018-11-16 发布日期:2018-11-16

Quantum Genetic Evolutionary Algorithm for Bin Packing Problem

GUO Jing and CHEN Xian-fu   

  • Online:2018-11-16 Published:2018-11-16

摘要: 针对遗传算法系统的维持能力问题,提出一种量子演化算法(a Quantum-Inspired Evolutionary Algorithm)用于解决装箱问题的布局与优化。算法中采用量子比特编码、量子延伸变异操作。同时根据装箱问题具体情况,设计相应的量子旋转门更新策略,并在此基础上引入遗传操作,同时提出MCBF算法修复策略。最后,对8个测试数据集进行测试。实验测试结果显示,算法在维持遗传基因种群多样性与提高优化质量等方面效果明显。

关键词: 量子演化,遗传操作,装箱问题,种群多样性

Abstract: Aiming at the problem of the genetic algorithm’s system maintenance ability,the paper presented a quantum-inspired evolutionary algorithm for layout and optimization of bin packing problem.This algorithm adopts quantum bit coding and quantum improving mutation.According to the specific situation of bin packing problem,the corresponding rotation gate updating strategy was devised and genetic operation was introduced.Furthermore a new repair strategy called modified complete and best fit algorithm were proposed.Finally,the eight testing sets were tested.The experimental results show this algorithm has a higher performance in maintaining the population diversity of genetic gene and improving the quality of optimization.

Key words: Quantum-inspired,Genetic operation,Bin packing problem,Population diversity

[1] 实例数目总和HGGAMGGAQIEA 最优值运行时间最优值运行时间最优值运行时间 u12020181.701292180.635905200.958405 u250201739.6108231813.803196.937677 u5002018752.0479418108.8121957371164 u100020166572.956217588.8807518585.845125 T6020620.789710.6953181.216115 T120200053.76911222.20842 Total120757387.10583726.5961106684.5369 测试硬件条件为:3.10GHz,3.49G的内存,WIN-XP操作系统,VC++环境。 表4中“最优值”表示该算法成功解决的实例数目,“运行时间”为每个实例每次的平均运行时间。 表4的实验数据显示,QIEA算法在解决数据集u120,u250,u500,u1000时,数据集中每个实例的平均运行时间都小于HGGA算法的平均运行时间,特别是当解决数据集u1000时,QIEA的平均运行时间更是远远小于HGGA算法。 MGGA算法有两个数据集的运行时间要优于QIEA算法,分析可能原因如下:由于QIEA算法引入了量子旋转门更新操作,提高了算法的空间复杂性和计算复杂性,特别是计算时涉及到了浮点运算,因此可能延长算法的执行时间。 由于QIEA算法利用了量子比特编码,因此提高了种群的多样性和系统维持能力,克服了遗传算法的早熟收敛的情况。同时,QIEA算法与HGGA算法相比较,其不需要提前计算所使用箱子数目的上限值,使算法更加简洁。 在解决数据集T60和T120时,QIEA算法的效果更加明显,不仅能成功解决更多的实例数目,而且运行时间更短。 结束语 理论分析与部分实验结果显示,在系统维持能力和种群多样性方面,QIEA算法优于HGGA算法和MGGA算法,面向特种装箱的实验应用方面可以显示良好的优化效果。需要指出的是装箱问题是一个强约束多目标优化经典NP难题,有关一般三维强约束装箱应用问题有待进一步扩展研究。 Beasley J E.OR-Library:Distributing test problems by electronicmail [J].Journal of Operational Research Society,1990,41(11):1069-1072 (下转第102页)(上接第69页)
[2] Singh A,Gupta A K.Two heuristics for the one-dimensionalbin-packing problem [J].OR Spectrum,2007,29(4):765-781
[3] Kuk-Hyun,Kim J-H.Quantum-Inspired Evolutionary Algo-rithm for a Class of Combinatorial Optimization [J].Evolutionary Computation,2002,6(6):580-593
[4] Gu Jin-wei,Gu Man-zhan.Anovel competitive co-evolutionaryquantum genetic algorithm for stochastic job shop scheduling problem[J].Computers and Operations Research,2010,7(5):927-937
[5] Conffman E G,Galambos J G.Bin packing approximation algorithms:combinatorial analysis [M].the Netherlands:Kluwer Academic Publishers,1998:151-207
[6] 杨俊安,庄镇全.量子遗传算法研究现状[J].计算机科学,2003,0(11):13-15
[7] Galinier P,Hao J K.Hybrid evolutionary algorithms for graph coloring[J].Journal of Combinatorial Optimization,1999,3(4):381-383
[8] Han K-H,Kim J-H.Genetic Quantum Algorithm and its Application to Combinatorial Optimization Problem [J].Evolutionary Computation,2000,2:1354-1360
[9] 。 表3 修复策略算法结果比较(使用箱子数目的平均值) u120u250u500u1000T60T120T249T502 NFD68.6142.6282.4562.724.648.6100.1201.2 FFD49.8103.1203.9405.423.245.895190.1 BFD49.8103.1203.9405.423.245.895190.1 MFFD49.7103203.6404.923.245.895190.1 CBF49.8103.1203.5405.12345.794.9190.4 MCBF50.8103.4204.5404.421.842.387.2173.9 表3的实验结果显示,MCBF算法除了数据集u120,u500以外,都能得到比其他算法更优的解,且在数据集T60,T120,T249,T501上,解的质量有优。所以本文采用MCBF算法作为QIEA算法的修复策略。 表4是HGGA算法和MGGA算法与本文提出的QIEA算法的实验结果比较。表中HGGA算法和MGGA算法分别运行100次。QIEA算法运行1次,且交叉概率pcross=0.8,量子变异概率pmutation=0.1。 表4 QIEA算法、MGGA算法和HGGA算法的测试比较结果表 数据 集[1]实例数目总和HGGAMGGAQIEA 最优值运行时间最优值运行时间最优值运行时间 u12020181.701292180.635905200.958405 u250201739.6108231813.803196.937677 u5002018752.0479418108.8121957371164 u100020166572.956217588.8807518585.845125 T6020620.789710.6953181.216115 T120200053.76911222.20842 Total120757387.10583726.5961106684.5369 测试硬件条件为:3.10GHz,3.49G的内存,WIN-XP操作系统,VC++环境。 表4中“最优值”表示该算法成功解决的实例数目,“运行时间”为每个实例每次的平均运行时间。 表4的实验数据显示,QIEA算法在解决数据集u120,u250,u500,u1000时,数据集中每个实例的平均运行时间都小于HGGA算法的平均运行时间,特别是当解决数据集u1000时,QIEA的平均运行时间更是远远小于HGGA算法。 MGGA算法有两个数据集的运行时间要优于QIEA算法,分析可能原因如下:由于QIEA算法引入了量子旋转门更新操作,提高了算法的空间复杂性和计算复杂性,特别是计算时涉及到了浮点运算,因此可能延长算法的执行时间。 由于QIEA算法利用了量子比特编码,因此提高了种群的多样性和系统维持能力,克服了遗传算法的早熟收敛的情况。同时,QIEA算法与HGGA算法相比较,其不需要提前计算所使用箱子数目的上限值,使算法更加简洁。 在解决数据集T60和T120时,QIEA算法的效果更加明显,不仅能成功解决更多的实例数目,而且运行时间更短。 结束语 理论分析与部分实验结果显示,在系统维持能力和种群多样性方面,QIEA算法优于HGGA算法和MGGA算法,面向特种装箱的实验应用方面可以显示良好的优化效果。需要指出的是装箱问题是一个强约束多目标优化经典NP难题,有关一般三维强约束装箱应用问题有待进一步扩展研究。 Beasley J E.OR-Library:Distributing test problems by electronicmail [J].Journal of Operational Research Society,1990,41(11):1069-1072 (下转第102页)(上接第69页)[2]Singh A,Gupta A K.Two heuristics for the one-dimensionalbin-packing problem [J].OR Spectrum,2007,29(4):765-781[3]Kuk-Hyun,Kim J-H.Quantum-Inspired Evolutionary Algo-rithm for a Class of Combinatorial Optimization [J].Evolutionary Computation,2002,6(6):580-593[4]Gu Jin-wei,Gu Man-zhan.Anovel competitive co-evolutionaryquantum genetic algorithm for stochastic job shop scheduling problem[J].Computers and Operations Research,2010,7(5):927-937[5]Conffman E G,Galambos J G.Bin packing approximation algorithms:combinatorial analysis [M].the Netherlands:Kluwer Academic Publishers,1998:151-207[6]杨俊安,庄镇全.量子遗传算法研究现状[J].计算机科学,2003,0(11):13-15[7]Galinier P,Hao J K.Hybrid evolutionary algorithms for graph coloring[J].Journal of Combinatorial Optimization,1999,3(4):381-383[8]Han K-H,Kim J-H.Genetic Quantum Algorithm and its Application to Combinatorial Optimization Problem [J].Evolutionary Computation,2000,2:1354-1360[9]Kim B-I,Wy J.Last two fit augmentation to the well-knownconstruction heuristics for one-dimensional bin-packing problem:an empirical study [J].The International Journal of Advanced Manufacturing Technology,2010,50(9-12):1145-1152
[10] Stawowy A.Evolutionary based heuristic for bin packing problem[J].Computer and Industrial Engineering,2008,5(2):465-474
[11] Bhatia A K,Basu S K.Packing Bins Using Multi-chromosomal Genetic Representation and Better-Fit Heuristic [J].Computer Science,2004,3316:181-186
[12] Chan F T S,Au K C,Chan L Y,et al.Using genetic algorithms to solve quality-related bin packing problem[J].Robotics and Computer-Integrated Manufacturing,2007,3(1):71-72
[13] 陈国良,等.遗传算法及其应用[M].北京:人民邮电出版社,2001:101-161
[14] Gupta J N D,Ho J C.A new heuristic algorithm for the one-dimensional bin-packing problem[J].Production planning & control,1999,0(6):598-603

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!