資源描述:
《改進(jìn)搜索策略的人工蜂群算法》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、JournalofComputerApplicationsISSN1001-90812012-12-01計(jì)算機(jī)應(yīng)用,2012,32(12):3326-3330CODENJYIIDUhttp://www.joca.cn文章編號(hào):1001-9081(2012)12-3326-05doi:10.3724/SP.J.1087.2012.03326改進(jìn)搜索策略的人工蜂群算法*張銀雪,田學(xué)民,曹玉蘋(中國(guó)石油大學(xué)(華東)信息與控制工程學(xué)院,山東青島266580)(*通信作者電子郵箱wendyzyx@yaho
2、o.cn)摘要:針對(duì)人工蜂群(ABC)算法存在收斂速度慢、收斂精度低的問(wèn)題,給出一種改進(jìn)的人工蜂群算法用于數(shù)值函數(shù)優(yōu)化問(wèn)題。在ABC的鄰域搜索公式中利用目標(biāo)函數(shù)自適應(yīng)調(diào)整步長(zhǎng),并根據(jù)迭代次數(shù)非線性減小偵查蜂的搜索范圍。改進(jìn)ABC算法提高了ABC算法的局部搜索能力,能夠有效避免早熟收斂?;?個(gè)標(biāo)準(zhǔn)測(cè)試函數(shù)的仿真實(shí)驗(yàn)表明,改進(jìn)ABC算法的尋優(yōu)能力有較大提高,對(duì)于多個(gè)高維多模態(tài)函數(shù)該算法可取得理論全局最優(yōu)解。與對(duì)比算法相比,該算法具有更高的收斂精度,并且收斂速度更快。關(guān)鍵詞:人工蜂群算法;數(shù)值函數(shù)優(yōu)
3、化;鄰域搜索;自適應(yīng);非線性函數(shù)中圖分類號(hào):TP18文獻(xiàn)標(biāo)志碼:AArtificialbeecolonyalgorithmwithmodifiedsearchstrategy*ZHANGYin-xue,TIANXue-min,CAOYu-ping(CollegeofInformationandControlEngineering,ChinaUniversityofPetroleum(EastChina),QingdaoShandong266580,China)Abstract:Amodified
4、ArtificialBeeColony(ABC)algorithmwasproposedfornumericalfunctionoptimizationinthispaper,inordertosolvetheproblemsofslowconvergenceandlowcomputationalprecisionofconventionalABCalgorithm.ThemodifiedABCalgorithmcanadjustthestepsizeoftheselectedneighborf
5、oodsourcepositionadaptivelyaccordingtotheobjectivefunction.Ontheotherhand,thesearchingmethodbasedonanonlinearadjustmentofsearchrangedependingontheiterationwasintroducedforscoutbees.ThemodifiedABCalgorithmcanimprovetheexploitation,andavoidstheprematur
6、econvergenceeffectively.Theexperimentalresultsonsixbenchmarkfunctionsshowthat,themodifiedABCalgorithmsignificantlyimprovestheoptimizationability.ThemodifiedABCalgorithmcanachievetheglobalminimumvaluesfornumerousmultimodalfunctionswithhighdimension.Co
7、mparedtotheotherapproaches,theproposedmethodnotonlyobtainshigherqualitysolutions,butalsohasafasterconvergencespeed.Keywords:ArtificialBeeColony(ABC)algorithm;numericalfunctionoptimization;neighborhoodsearching;adaptive;nonlinearfunction因此,為改善ABC算法的全局
8、收斂性能,提高收斂速度,許多0引言[9-14][11]學(xué)者對(duì)算法進(jìn)行了改進(jìn)。例如,Banharnsakun等提出函數(shù)優(yōu)化問(wèn)題在科學(xué)實(shí)驗(yàn)、工程設(shè)計(jì)和生產(chǎn)實(shí)踐等方面利用當(dāng)前全局最優(yōu)解代替隨機(jī)選取的鄰域個(gè)體,并根據(jù)當(dāng)前有著廣泛應(yīng)用。近年來(lái),群智能優(yōu)化算法以其較好的通用性、全局最優(yōu)解的適應(yīng)度調(diào)整鄰域搜索步長(zhǎng),提高收斂精度及收[12]容錯(cuò)性及對(duì)初始值不敏感等優(yōu)點(diǎn)成為解決函數(shù)優(yōu)化問(wèn)題的一斂速度。Li等提出利用全局最優(yōu)解、慣性權(quán)重和加速系數(shù)[1]種新途徑。人工蜂群(ArtificialBeeColony,AB