广西师范学院学报(自然科学版)
廣西師範學院學報(自然科學版)
엄서사범학원학보(자연과학판)
JOURNAL OF GUANGXI TEACHERS EDUCATION UNIVERSITY(NATURAL SCIENCE EDITION)
2014年
1期
50-55
,共6页
元建%覃晓%彭昱忠%石亚冰
元建%覃曉%彭昱忠%石亞冰
원건%담효%팽욱충%석아빙
多细胞%基因表达式编程%多核处理器%并行算法
多細胞%基因錶達式編程%多覈處理器%併行算法
다세포%기인표체식편정%다핵처리기%병행산법
multicellular%gene expression programming%multicore processors%parallel algo-rithm
并行算法是当前研究解决算法效率问题的成熟技术之一。为提高GEP算法解决复杂函数优化问题的效率,将并行算法引入多细胞基因表达式编程函数优化问题,解决传统计算形式不能充分发挥多核处理器性能的问题。通过分析多细胞基因表达式编程并行算法的机理和M PI和O pen M P混合并行模型,设计与实现多细胞基因表达式编程函数优化的并行算法(Parallel Multicellular Gene Expression Programming algorithm for Function Optimization) PGMFO。实验结果表明针对复杂的函数优化问题,在不影响精度和收敛性的情况下,PGMFO算法比原有的算法效率高出10%~20%。
併行算法是噹前研究解決算法效率問題的成熟技術之一。為提高GEP算法解決複雜函數優化問題的效率,將併行算法引入多細胞基因錶達式編程函數優化問題,解決傳統計算形式不能充分髮揮多覈處理器性能的問題。通過分析多細胞基因錶達式編程併行算法的機理和M PI和O pen M P混閤併行模型,設計與實現多細胞基因錶達式編程函數優化的併行算法(Parallel Multicellular Gene Expression Programming algorithm for Function Optimization) PGMFO。實驗結果錶明針對複雜的函數優化問題,在不影響精度和收斂性的情況下,PGMFO算法比原有的算法效率高齣10%~20%。
병행산법시당전연구해결산법효솔문제적성숙기술지일。위제고GEP산법해결복잡함수우화문제적효솔,장병행산법인입다세포기인표체식편정함수우화문제,해결전통계산형식불능충분발휘다핵처리기성능적문제。통과분석다세포기인표체식편정병행산법적궤리화M PI화O pen M P혼합병행모형,설계여실현다세포기인표체식편정함수우화적병행산법(Parallel Multicellular Gene Expression Programming algorithm for Function Optimization) PGMFO。실험결과표명침대복잡적함수우화문제,재불영향정도화수렴성적정황하,PGMFO산법비원유적산법효솔고출10%~20%。
T he parallel algorithm is currently a mature technology to solve the efficiency of algo-rithm .This article introduces the idea of parallel algorithms into multicellular gene expression pro-gramming algorithm for function optimization to solve the shortage that traditional computing form can not give full play to the performance of the current multi-core processors for improving the effi-ciency of GEP algorithm in solving complex function optimization problems .By analyzing the mecha-nism of multicellular gene expression programming algorithm for function optimization and the mixed parallel model of MPI and Open MP ,In this paper a parallel multicellular gene expression program-ming algorithm for function optimization is designed and achieved .The experimental results show that in the case of not affecting the accuracy and convergence ,and in view of the complex function optimi-zation ,the efficiency of PGM FO algorithm improve by 10% ~ 20% .