Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (12): 77-80.
• 学术探讨 • Previous Articles Next Articles
Received:
Revised:
Online:
Published:
杨广益 欧阳智敏 全惠云
通讯作者:
Abstract: Evolutionary Computation belongs to the algorithms inspired by nature. Estimation of Distribution Algorithms (EDAs) is based on the simulation and inference of probability distribution of population. Practices in reference[1] have showed the advantage of EDAs. Inspired by the complementarity mechanism in nature, this paper present a Relaxed Complemental Estimation of Distribution Algorithm (RCEDA). We carry out experiment studies on the well-know Chu and Beasley MPK benchmark. The analysis and computational results show that our algorithm is competitive
Key words: Linear Programming, Multidimesional Knapsack Problem, Estimation of Distribution Algorithms, Relaxed complemental model.
摘要: 演化计算(Evolutionary Computation 简记为EC) 是受自然界物种进化启发而产生的一类优化技术。分布估计算法(Estimation of Distribution Algorithms 简记为EDAs)是一类基于概率模型的估计与模拟的演化算法,它在参数设置及计算效率上有着其它演化方法难以比拟的优势[1]。受自然界互补机制的启发,本文提出了一种改进的分布估计算法-松驰互补分布估计算法(Relaxed Complemental Estimation of Distribution Algorithm 简记为RCEDA)。在求解Chu 和 Beasley提出的MKP Benchmark[3] 时,算法在较短的时间内找到了大多数目前书籍的最好解,分析和实验结果表明RCEDA是一种较好的演化算法。
关键词: 线性规划, 多维背包问题, 分布估计算法, 松驰互补模型
杨广益 欧阳智敏 全惠云. 松驰互补分布估计算法求解多维背包问题[J]. 计算机工程与应用, 2007, 43(12): 77-80.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://cea.ceaj.org/EN/
http://cea.ceaj.org/EN/Y2007/V43/I12/77