计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2010年
11期
29-31,45
,共4页
子群%粒子群算法%郭涛算法
子群%粒子群算法%郭濤算法
자군%입자군산법%곽도산법
subgroup%Particle Swarm Optimization(PSO) algorithm%Guotao algorithm
从研究分析粒子群算法和郭涛算法的特点出发,提出一种综合两算法优点的混合算法.新算法改变了粒子的更新方式,以子空间搜索和串行搜索相结合的多点并行搜索,扩大了算法的搜索范围,减少了粒子对初值的依赖,增强了算法跳出局部最优的能力;通过后代较优个体变异产生子群,提高了算法局部寻优能力;实验证明,该算法正确高效.
從研究分析粒子群算法和郭濤算法的特點齣髮,提齣一種綜閤兩算法優點的混閤算法.新算法改變瞭粒子的更新方式,以子空間搜索和串行搜索相結閤的多點併行搜索,擴大瞭算法的搜索範圍,減少瞭粒子對初值的依賴,增彊瞭算法跳齣跼部最優的能力;通過後代較優箇體變異產生子群,提高瞭算法跼部尋優能力;實驗證明,該算法正確高效.
종연구분석입자군산법화곽도산법적특점출발,제출일충종합량산법우점적혼합산법.신산법개변료입자적경신방식,이자공간수색화천행수색상결합적다점병행수색,확대료산법적수색범위,감소료입자대초치적의뢰,증강료산법도출국부최우적능력;통과후대교우개체변이산생자군,제고료산법국부심우능력;실험증명,해산법정학고효.
This essay starts with the analysis of the characteristics of particle swarm optimization and Guotao algorithm,and raises a hybrid algorithm with their advantages.The new algorithm changes the particles' updating ways,searches with the combination of subspace and serial searches,enlarges the algorithm searching scope,reduces the particle's dependency to the initial value, strengthens its ability out of the partial optimalization.lt also produces subgroups by the better individuals' variation of the descendants,improves its partial optimization ability.Experiments show that the new algorithm is high-efficiency.