新型果蠅優(yōu)化算法的研究
發(fā)布時間:2018-12-12 02:10
【摘要】:由于傳統(tǒng)果蠅優(yōu)化算法(FOA)具有尋優(yōu)精度低和易陷入局部極小點的缺點,提出了一種具有不同飛行半徑的分群搜索策略,使得在搜索區(qū)間內(nèi)果蠅的種群多樣性大大增加;同時在果蠅個體的飛行距離與方向的步長函數(shù)上,針對不同的果蠅子群引入了不同的函數(shù),該類函數(shù)具有周期震蕩性質(zhì),可以很好地避免果蠅群陷入局部極小點而無法求得最優(yōu)解。通過對8個測試函數(shù)的仿真實驗,驗證了這些策略能夠有效地提高搜索精度、收斂速度和穩(wěn)定性。
[Abstract]:Because the traditional Drosophila optimization algorithm (FOA) has the disadvantages of low precision and easy to fall into local minima, a cluster search strategy with different flight radius is proposed, which greatly increases the population diversity of Drosophila in the search region. At the same time, different functions are introduced for different Drosophila subgroups on the step function of flying distance and direction of Drosophila, which has the property of periodic oscillation. It can avoid the drosophila population falling into local minima and can not get the optimal solution. The simulation results of eight test functions show that these strategies can effectively improve the search accuracy, convergence speed and stability.
【作者單位】: 安徽大學計算智能與信號處理重點實驗室;安徽大學計算機科學與技術學院;
【基金】:安徽省科技攻關項目(No.06060701)
【分類號】:TP18
本文編號:2373695
[Abstract]:Because the traditional Drosophila optimization algorithm (FOA) has the disadvantages of low precision and easy to fall into local minima, a cluster search strategy with different flight radius is proposed, which greatly increases the population diversity of Drosophila in the search region. At the same time, different functions are introduced for different Drosophila subgroups on the step function of flying distance and direction of Drosophila, which has the property of periodic oscillation. It can avoid the drosophila population falling into local minima and can not get the optimal solution. The simulation results of eight test functions show that these strategies can effectively improve the search accuracy, convergence speed and stability.
【作者單位】: 安徽大學計算智能與信號處理重點實驗室;安徽大學計算機科學與技術學院;
【基金】:安徽省科技攻關項目(No.06060701)
【分類號】:TP18
【參考文獻】
相關期刊論文 前7條
1 武興宇;孫磊;胡翠云;孫瑞辰;;基于改進粒子群優(yōu)化算法的虛擬機遷移選擇策略研究[J];計算機科學;2015年S1期
2 寧桂英;周永權;;基于差分進化算法的收斂性分析[J];南通大學學報(自然科學版);2014年03期
3 吳小文;李擎;;果蠅算法和5種群智能算法的尋優(yōu)性能研究[J];火力與指揮控制;2013年04期
4 韓俊英;劉成忠;;基于細菌趨化的果蠅優(yōu)化算法[J];計算機應用;2013年04期
5 潘文超;;應用果蠅優(yōu)化算法優(yōu)化廣義回歸神經(jīng)網(wǎng)絡進行企業(yè)經(jīng)營績效評估[J];太原理工大學學報(社會科學版);2011年04期
6 王聯(lián)國;洪毅;趙付青;余冬梅;;一種改進的人工魚群算法[J];計算機工程;2008年19期
7 畢惟紅;任紅民;吳慶標;;一種新的遺傳算法最優(yōu)保存策略[J];浙江大學學報(理學版);2006年01期
,本文編號:2373695
本文鏈接:http://sikaile.net/kejilunwen/zidonghuakongzhilunwen/2373695.html
最近更新
教材專著