摘要
首先针对演化算法求解背包问题定义了贪心变换的概念,并给出了该变换的一种有效实现算法;然后将此算法与文献[5]中提出的具有双重结构编码的二进制粒子群优化算法(DS_BPSO)相结合,提出了一种解决广义背包问题GKP(General Knapsack Problem)的快速算法:基于贪心变换的DS_BPSO算法(GDS_BPSO)。利用该算法求解文献[3,6]中的著名背包实例,给出了该背包实例的目前最好结果。此外,对于随机生成的大规模背包实例,通过与文献[3]中的HGA算法对比计算表明:GDS_BPSO算法是求解广义背包问题的一种高效方法。
Firstly, the concept of greedy transformation is defined according to the evolution algorithm for knapsack problem, and an effective algorithm is proposed to implement greedy transformation. Then, through the combination of this algorithm and binary particle swarm optimization with double-structure coding given in [ 5 ] , solve general knapsack problem （GKP） , a rapid DS BPSO algorithm based on greedy transformation （ GDS BPSO） is advanced to solve general knapsack problem（GKP）. GDS BPSO is applied ＂to solve the knapsack sample in [ 3,6 ], and the best result of the sample is given. Through the contrast to HGA algorithm in [ 3 ] for random generated large-scale knapsack samples,it is showed that GDS BPSO algorithm is an effective solution to general knapsack problem.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第4期230-232,262,共4页
Computer Applications and Software
作者简介
贺毅朝，副教授，主研领域：智能计算，计算机密码学和算法理论。
