计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2014年
1期
58-62
,共5页
粒子群优化算法%遗传算法%伙伴选择
粒子群優化算法%遺傳算法%夥伴選擇
입자군우화산법%유전산법%화반선택
particle swam optimization%genetic algorithm%partner selection
针对标准粒子群优化算法易陷入局部最优的缺点,提出了一种遗传粒子群混合算法。通过对算法中惰性粒子和局部最优粒子分别进行交叉变异,以及消除粒子速度对寻优的干扰,从而避免了粒子种群单一化和局部最优的问题。将该算法应用于虚拟企业伙伴选择实验,结果表明在进化代数和最优值方面是令人满意的。
針對標準粒子群優化算法易陷入跼部最優的缺點,提齣瞭一種遺傳粒子群混閤算法。通過對算法中惰性粒子和跼部最優粒子分彆進行交扠變異,以及消除粒子速度對尋優的榦擾,從而避免瞭粒子種群單一化和跼部最優的問題。將該算法應用于虛擬企業夥伴選擇實驗,結果錶明在進化代數和最優值方麵是令人滿意的。
침대표준입자군우화산법역함입국부최우적결점,제출료일충유전입자군혼합산법。통과대산법중타성입자화국부최우입자분별진행교차변이,이급소제입자속도대심우적간우,종이피면료입자충군단일화화국부최우적문제。장해산법응용우허의기업화반선택실험,결과표명재진화대수화최우치방면시령인만의적。
For the standard particle swarm optimization algorithm is easy to fall into local extremum, this paper proposes a genetic and particle swarm optimization hybrid algorithm. The hybrid algorithm has avoided the problems of single par-ticle swam and local extremum by executing crossover and mutation operation to the lazy particles and the local optimum particles, eliminating the interference of the particle velocity to the optimization process. Finally, this algorithm has been applied to the experiment of virtual enterprise partner selection, the experiment show that it has satisfied results in evolu-tion generations and optimal value.