資源描述:
《基于改進(jìn)蜂群算法的冷連軋規(guī)程優(yōu)化設(shè)計(jì).pdf》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫。
1、第34卷第5期礦冶工程V0l_34№52014年10月MININGANDMETALLURGICALENGINEERING0ctober2014基于改進(jìn)蜂群算法的冷連軋規(guī)程優(yōu)化設(shè)計(jì)①魏立新,呂白,李瑩,楊景明’(1.國家冷軋板帶裝備及工藝工程技術(shù)研究中心,河北秦皇島066004;2.燕山大學(xué)工業(yè)計(jì)算機(jī)控制河北省重點(diǎn)實(shí)驗(yàn)室,河北秦皇島066004)摘要:綜合考慮現(xiàn)場(chǎng)和設(shè)備所受的約束條件,以等負(fù)荷和克服打滑為目標(biāo)函數(shù),建立了軋制規(guī)程多目標(biāo)優(yōu)化模型。為了提高算法性能,對(duì)人工蜂群算法進(jìn)行了改進(jìn)。首先,應(yīng)用反向?qū)W習(xí)的策略初始化種群,使得個(gè)
2、體盡可能均勻分布在搜索空間。其次,人工蜂群算法采用不同的選擇機(jī)制,提高收斂速度和尋優(yōu)精度。最后,用改進(jìn)的算法對(duì)某五機(jī)架冷連軋機(jī)進(jìn)行規(guī)程優(yōu)化設(shè)計(jì)。結(jié)果表明,改進(jìn)的人工蜂群算法能有效避免早熟收斂,全局優(yōu)化能力和收斂速率都有顯著提高。關(guān)鍵詞:規(guī)程優(yōu)化;人工蜂群算法;反向?qū)W習(xí);選擇策略中圖分類號(hào):TG339文獻(xiàn)標(biāo)識(shí)碼:Adoi:10.3969/j.issn.0253—6099.2014.05.029文章編號(hào):0253-6099(2014)05-0118-05OptimizationDesignofRollingScheduleforT
3、andemColdMiUBasedonModifiedArtifidalBeeColonyAlgorithmWEILi—xin’,LUBai,LIring,rANGJing-ming’(1.NationalEngineeringResearchCenterforEquipmentandTechnologyofColdStripRolling,Qinhuangdao066004,Hebei,China;2.KeyLabofIndustrialComputerControlEngineeringofHebeiProvince,Ya
4、nshanUniversity,Qinhuangdao066004,Hebei,China)Abstract:Withcertainconstraintconditionsoffacilitiesonengineeringsitetakenintoconsideration.a(chǎn)muhiobjectiveoptimizationmodelforrollingschedulewasestablishedwithequalizingrollingloadandovercomingslippageasobjectivefunction
5、s.Theartificialbeecolony(ABC)algorithmwasmodifiedtoimproveitsperformance.Firstly,aninitializationstrategybasedontheopposition—basedlearningwasappliedtodiversifyhomogeneouslytheindividualsinthesearchspace.Then,severalselectionstrategieswereappliedthroughsimulationtoi
6、mprovetheoptimizingaccuracyandacceleratetheconvergence.Finally,scheduleoptimizationstrategyforafive—standtandemrollingmillWasdesignedbasedOfthemodifiedalgorithm.Theresultsdemonstratethat,themodifiedalgorithmcannotonlyavoideffectivelytheprematureconvergence,butalsoim
7、provetheoverall—optimizationabilityandtheconvergencespeed.Keywords:scheduleoptimization;artificialbeecolony(ABC);algorithmopposition—basedlearning;selectionstrategy合理的軋制規(guī)程既可提高冷軋帶鋼的生產(chǎn)率,又究中,與遺傳算法、微粒群等其他優(yōu)化算法相比,人工能保證產(chǎn)品質(zhì)量,提高工藝控制的精度和響應(yīng)速度以蜂群算法的突出優(yōu)點(diǎn)是每次迭代都進(jìn)行全局和局部搜及設(shè)備的利用效率,帶來
8、極大的經(jīng)濟(jì)效益¨j。在實(shí)索,找到最優(yōu)解的概率大大增加,并在較大程度上避免際生產(chǎn)中帶鋼與工作輥表面經(jīng)常會(huì)出現(xiàn)打滑現(xiàn)象。尤了局部最優(yōu)J。本文首先建立了一套以等負(fù)荷和預(yù)其是在高速生產(chǎn)時(shí),這種現(xiàn)象尤為突出]。因此,確防打滑為目標(biāo)函數(shù)的冷連軋機(jī)軋制規(guī)程的綜合優(yōu)化模定合理的軋制規(guī)程,克服