資源描述:
《基于蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡的數(shù)控機床熱誤差補償基于蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡的數(shù)控機床熱誤差補償.pdf》由會員上傳分享,免費在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫。
1、第10期組合機床與自動化加工技術(shù)N0.102013年10月ModularMachineTool&AutomaticManufacturingTechniqueoct.2013文章編號:1001—2265(2013)10—0050—04基于蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡的數(shù)控機床熱誤差補償章曉英(上海行健職業(yè)學院信息技術(shù)與機電工程系,上海200072)摘要:文章針對影響數(shù)控機床加工精度的溫度誤差進行研究分析,提出了一種基于蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡的數(shù)控機床熱誤差補償方法。文章詳細闡述了蟻群算法和BP神經(jīng)網(wǎng)絡算法,以及
2、蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡過程,給出了數(shù)控機床熱誤差補償硬件系統(tǒng)。以三軸聯(lián)動臥式加工中心為例,在合理布置熱傳感器的同時,利用粗集理論相關(guān)知識方法,提取機床熱誤差補償重要特征參數(shù)。運用蟻群算法優(yōu)化BP神經(jīng)網(wǎng)絡建立精確較高和魯棒性較強的熱誤差模型。通過實驗仿真結(jié)果表明經(jīng)補償機床加工精度得到較大提升,補償效果也明顯優(yōu)于BP神經(jīng)網(wǎng)絡以及PSO.BP神經(jīng)網(wǎng)絡,具有一定的理論意義和實用價值。關(guān)鍵詞:蟻群算法;BP神經(jīng)網(wǎng)絡;熱誤差補償;數(shù)控機床中圖分類號:TH161文獻標識碼:AThermalErrorCompensatio
3、nforCNCMachineToolsBasedonACO.BPNeuralNetworkZHANGXiao—ying(DepartmentofInformationTechnologyandMechatronicsEngineering,ShanghaiXingjianCollege,Shang-hai200072,China)Abstract:FocusingonthestudyandanalysisonofMachiningaccuracyofNCmachinetoolsthermaler—ror,th
4、ispaperhaspresentedaNCmachinetoolsthermalerrorcompensationmethodbasedonAntColo—nyoptimizationalgorithmBPneuralnetwork.ThisarticlehaselaboratedontheAntColonyalgorithm,BPneuralnetworkalgorithmandAntColonyoptimizationalgorithmBPneuralnetwork,andhasgiventhehard
5、waresystemofNCmachinetoo1therma1errorcompensation.Withanexampleofthree.a(chǎn)xishori.zontalmachiningcenter,atthesametimeofreasonablelayoutofthermalsensors,usingroughsettheoryandknowledge,importantcharacteristicparametersofthermalerrorcompensationonmachinetoolsha
6、sbeenextracted.UsingAntColonyoptimizationalgorithmBPneuralnetwork,athermalerrormodelwithhigherprecisionandstrongrobustnesshasbeenbuilt.Throughthesimulatingexperimentalsimulation,itisshowedthattheappliedtothemachiningaccuracyhasbeengreatlyimprovedafterusingc
7、ompensation,andthecompensationeffectissignificantlybetterthanBPneuralnetworkaswellasthePSO—BPneuralnetwork,SOthismethodhassometheoreticalsignificanceandpracticalvalue.Keywords:antcolonyalgorithm;BPneuralnetwork;thermalerrorcompensation;CNCmachinetools照和環(huán)境溫度
8、,內(nèi)熱源的種類有很多,如電機傳動過程0引言中的摩擦熱、機械機構(gòu)操作產(chǎn)生的熱、切削過程的切削隨著精密加工技術(shù)的廣泛使用和發(fā)展,在很大程熱、軸承轉(zhuǎn)動過程中的摩擦熱等。。同時,機床零件度上提高了數(shù)控機床的加工精度。一般而言影響精密的復雜形狀、材料的熱特性差異等不確定因素使得機加工機床的誤差主要來自熱誤差,大概占總誤差的床的產(chǎn)熱機理非常復雜,所有這些因素都將導致機床40%~70%。針對如何減少熱誤差對機械加