HIDDEN-ARTICULATOR MARKOV MODELS PERFORMANCE IMPROVEMENTS AND ROBUSTNESS TO NOISE
- 格式:pdf
- 大小:38.18 KB
- 文档页数:4
HMM(隐马尔可夫模型)和CRF(条件随机场)是两种常用的概率图模型,它们都可以用来对序列数据进行建模,但是它们建模的对象不同。
HMM是一种生成式模型,它对联合概率分布进行建模。
在HMM中,我们假设存在一个隐藏的马尔可夫链,它生成了观测序列。
因此,HMM建模的是隐藏状态和观测序列的联合概率分布。
具体地,HMM通过定义状态转移概率和发射概率来描述隐藏状态和观测序列之间的关系,然后使用这些概率来计算联合概率分布。
相比之下,CRF是一种判别式模型,它对条件概率分布进行建模。
在CRF中,我们假设存在一个输入序列和一个对应的输出序列,我们想要建模的是给定输入序列下输出序列的条件概率分布。
具体地,CRF通过定义特征函数和对应的权重来描述输入序列和输出序列之间的关系,然后使用这些特征函数和权重来计算条件概率分布。
因此,HMM和CRF的主要区别在于它们建模的对象不同:HMM建模的是联合概率分布,而CRF建模的是条件概率分布。
这也导致了它们在应用上的不同:HMM常用于生成式任务,如语音识别、自然语言生成等;而CRF常用于判别式任务,如命名实体识别、分词等。
第9卷㊀第2期2024年3月气体物理PHYSICSOFGASESVol.9㊀No.2Mar.2024㊀㊀DOI:10.19527/j.cnki.2096 ̄1642.1098Mach数和壁面温度对HyTRV边界层转捩的影响章录兴ꎬ㊀王光学ꎬ㊀杜㊀磊ꎬ㊀余发源ꎬ㊀张怀宝(中山大学航空航天学院ꎬ广东深圳518107)EffectsofMachNumberandWallTemperatureonHyTRVBoundaryLayerTransitionZHANGLuxingꎬ㊀WANGGuangxueꎬ㊀DULeiꎬ㊀YUFayuanꎬ㊀ZHANGHuaibao(SchoolofAeronauticsandAstronauticsꎬSunYat ̄senUniversityꎬShenzhen518107ꎬChina)摘㊀要:典型的高超声速飞行器流场存在着复杂的转捩现象ꎬ其对飞行器的性能有着显著的影响ꎮ针对HyTRV这款接近真实高超声速飞行器的升力体模型ꎬ采用数值模拟方法ꎬ研究Mach数和壁面温度对HyTRV转捩的影响规律ꎮ采用课题组自研软件开展数值计算ꎬMach数的范围为3~8ꎬ壁面温度的范围为150~900Kꎮ首先对γ ̄Re~θt转捩模型和SST湍流模型进行了高超声速修正:将压力梯度系数修正㊁高速横流修正引入到γ ̄Re~θt转捩模型ꎬ并对SST湍流模型闭合系数β∗和β进行可压缩修正ꎻ然后开展了网格无关性验证ꎬ通过与实验结果对比ꎬ确认了修正后的数值方法和软件平台ꎻ最终开展Mach数和壁面温度对HyTRV边界层转捩规律的影响研究ꎮ计算结果表明ꎬ转捩区域主要集中在上表面两侧㊁下表面中心线两侧ꎻ增大来流Mach数ꎬ上下表面转捩起始位置均大幅后移ꎬ湍流区大幅缩小ꎬ但仍会存在ꎬ同时上表面层流区摩阻系数不断增大ꎬ下表面湍流区摩阻系数不断减小ꎻ升高壁面温度ꎬ上下表面转捩起始位置先前移ꎬ然后快速后移ꎬ最终湍流区先后几乎消失ꎮ关键词:转捩ꎻHyTRVꎻ摩阻ꎻMach数ꎻ壁面温度㊀㊀㊀收稿日期:2023 ̄12 ̄13ꎻ修回日期:2024 ̄01 ̄02基金项目:国家重大项目(GJXM92579)ꎻ广东省自然科学基金-面上项目(2023A1515010036)ꎻ中山大学中央高校基本科研业务费专项资金(22qntd0705)第一作者简介:章录兴(1998 )㊀男ꎬ硕士ꎬ主要研究方向为高超声速空气动力学ꎮE ̄mail:184****8082@163.com通信作者简介:张怀宝(1985 )㊀男ꎬ副教授ꎬ主要研究方向为空气动力学ꎮE ̄mail:zhanghb28@mail.sysu.edu.cn中图分类号:V211ꎻV411㊀㊀文献标志码:AAbstract:Thereisacomplextransitionphenomenonintheflowfieldofatypicalhypersonicvehicleꎬwhichhasasignifi ̄cantimpactontheperformanceofthevehicle.TheeffectsofMachnumberandwalltemperatureonthetransitionofHyTRVwerestudiedbynumericalsimulationmethods.Theself ̄developedsoftwareoftheresearchgroupwasusedtocarryoutnu ̄mericalcalculations.TherangeofMachnumberwas3~8ꎬandtherangeofwalltemperaturewas150~900K.Firstlyꎬthehypersoniccorrectionsoftheγ ̄Re~θttransitionmodelandtheSSTturbulencemodelwerecarriedout.Thepressuregradientcoefficientcorrectionandthehigh ̄speedcross ̄flowcorrectionwereintroducedintotheγ ̄Re~θttransitionmodelꎬandthecom ̄pressibilitycorrectionsoftheclosurecoefficientsβ∗andβoftheSSTturbulencemodelwerecarriedout.Thenꎬthegridin ̄dependenceverificationwascarriedoutꎬandthemodifiednumericalmethodandsoftwareplatformwereconfirmedbycom ̄paringwithexperimentalresults.FinallyꎬtheeffectsofMachnumberandwalltemperatureonthetransitionlawoftheHyTRVboundarylayerwerestudied.Theresultsshowthatthetransitionareaismainlyconcentratedonbothsidesoftheuppersurfaceandthecenterlineofthelowersurface.WiththeincreaseoftheincomingMachnumberꎬthestartingpositionoftransitionontheupperandlowersurfacesisgreatlybackwardꎬandtheturbulentzoneisgreatlyreducedꎬbutitstillex ̄ists.Atthesametimeꎬthefrictioncoefficientofthelaminarflowzoneontheuppersurfaceincreasescontinuouslyꎬandthefrictioncoefficientoftheturbulentzoneonthelowersurfacedecreases.Asthewalltemperatureincreasesꎬthestartingposi ̄tionoftransitionontheupperandlowersurfacesshiftsforwardꎬthenrapidlyshiftsbackwardꎬandfinallytheturbulentzonealmostdisappears.气体物理2024年㊀第9卷Keywords:transitionꎻHyTRVꎻfrictionꎻMachnumberꎻwalltemperature引㊀言高超声速飞行器具有突防能力强㊁打击范围广㊁响应迅速等显著优势ꎬ正逐渐成为各国空天竞争的热点[1]ꎮ高超声速飞行器边界层转捩是该类飞行器气动设计中的重要问题[2]ꎮ在边界层转捩过程中ꎬ流态由层流转变为湍流ꎬ飞行器的表面摩阻急剧增大到层流时的3~5倍ꎬ严重影响飞行器的气动性能与热防护系统ꎬ转捩还会导致飞行器壁面烧蚀㊁颤振加剧㊁飞行姿态控制难度大等一系列问题ꎬ对飞行器的飞行安全构成严重的威胁[3 ̄5]ꎬ开展高超声速飞行器边界层转捩研究具有十分重要的意义ꎮ影响边界层转捩的因素很多ꎬ例如ꎬMach数㊁Reynolds数㊁湍流强度㊁表面传导热等ꎮ在高超声速流动条件下ꎬ强激波㊁强逆压梯度㊁熵层等高超声速现象及其相互作用ꎬ会使得转捩流动的预测和研究难度进一步增大[6]ꎮ目前高超声速飞行器转捩数值模拟方法主要有直接数值模拟(DNS)㊁大涡模拟(LES)和基于Reynolds平均Navier ̄Stokes(RANS)的转捩模型方法ꎬ由于前两种计算量巨大ꎬ难以推广到工程应用ꎬ基于Reynolds平均Navier ̄Stokes的转捩模型在工程实践中应用最为广泛ꎬ其中γ ̄Re~θt转捩模型基于局部变量ꎬ与现代CFD方法良好兼容ꎬ目前已经有多项研究尝试从一般性的流动问题拓展到高超声速流动转捩模拟[6 ̄9]ꎮ目前高超声速流动转捩的研究对象主要是结构相对简单的构型ꎮMcDaniel等[10]研究了扩口直锥在高超声速流动条件下的转捩现象ꎮPapp等[11]研究了圆锥在高超声速流动条件下的转捩特性ꎮ美国和澳大利亚组织联合实施的HIFiRE计划[12]ꎬ研究了圆锥形状的HIFiRE1和椭圆锥形的HIFiRE5的转捩问题ꎮ杨云军等[13]采用数值模拟方法ꎬ分析了椭圆锥的转捩影响机制ꎬ并研究了Reynolds数对转捩特性的影响规律ꎮ另外ꎬ袁先旭等[14]于2015年成功实施了圆锥体MF ̄1航天模型飞行试验ꎮ以上对高超声速流动的转捩研究ꎬ都取得了比较理想的结果ꎬ然而所采用的模型都是圆锥㊁椭圆锥等简单几何外形ꎬ这与真实高超声速飞行器有较大差异ꎬ较难反映真实的转捩特性ꎮ为了有效促进对真实高超声速飞行器的转捩问题研究ꎬ中国空气动力研究与发展中心提出并设计了一款接近真实飞行器的升力体模型ꎬ即高超声速转捩研究飞行器(hypersonictransitionresearchvehicleꎬHyTRV)[15]ꎬ模型详细的参数见参考文献[16]ꎮHyTRV外形如图1所示ꎬ其整体外形较为复杂ꎬ不同区域发生转捩的情况也不尽相同ꎮ对HyTRV的转捩问题研究能够显著提高对真实高超声速飞行器转捩特性的认识水平ꎮLiu等[17]采用理论分析㊁数值模拟和风洞实验3种方法对HyTRV的转捩特性进行了研究ꎻ陈坚强等[15]分析了HyTRV的边界层失稳特征ꎻChen等[18]对HyTRV进行了多维线性稳定性分析ꎻQi等[19]在来流Mach数6㊁攻角0ʎ的条件下对HyTRV进行了直接数值模拟ꎻ万兵兵等[20]结合风洞实验与飞行试验ꎬ利用eN方法预测了HyTRV升力体横流区的转捩阵面形状ꎮ目前ꎬ相关研究主要集中在HyTRV的稳定性特征及转捩预测两个方面ꎬ而对若干关键参数ꎬ特别是Mach数和壁面温度对转捩的影响研究还比较少ꎮ(a)Frontview(b)Sideview㊀㊀㊀图1㊀HyTRV外形Fig.1㊀ShapeofHyTRV基于此ꎬ本文采用数值模拟方法ꎬ应用课题组自研软件开展Mach数和壁面温度对HyTRV转捩流动的影响规律研究ꎮ1㊀数值方法1.1㊀控制方程和数值方法控制方程为三维可压缩RANS方程ꎬ采用结构网格技术和有限体积方法ꎬ变量插值方法采用2阶MUSCL格式ꎬ通量计算采用低耗散的通量向量差分Roe格式ꎬ黏性项离散采用中心格式ꎬ时间推进方法采用LU ̄SGS格式ꎮ壁面采用等温㊁无滑移壁面条件ꎬ入口采用Riemann远场边界条件ꎬ出口采用零梯度外推边界条件ꎮ1.2㊀γ ̄Re~θt转捩模型γ ̄Re~θt转捩模型是Menter等[21ꎬ22]于2004年提01第2期章录兴ꎬ等:Mach数和壁面温度对HyTRV边界层转捩的影响出的一种基于拟合公式的间歇因子转捩模型ꎬ在2009年公布了完整的拟合公式及相关参数[23]ꎮ许多学者也开发了相应的程序ꎬ并进行了大量的算例验证[24 ̄28]ꎬ证明了该模型具有较好的转捩预测能力ꎬ预测精度较高ꎻ通过合适的标定ꎬγ ̄Re~θt转捩模型可以适用于多种情况下的转捩模拟ꎮ该模型构建了关于间歇因子γ的输运方程和关于转捩动量厚度Reynolds数Re~θt的输运方程ꎮ具体来说ꎬγ表示该位置是湍流流动的概率ꎬ取值范围为0<γ<1ꎮ关于γ的控制方程为Ə(ργ)Ət+Ə(ρujγ)Əxj=Pγ-Eγ+ƏƏxjμ+μtσfæèçöø÷ƏγƏxjéëêêùûúú其中ꎬPγ为生成项ꎬEγ为破坏项ꎮ关于Re~θt的输运方程为Ə(ρRe~θt)Ət+Ə(ρujRe~θt)Əxj=Pθt+ƏƏxjσθt(μ+μt)ƏRe~θtƏxjéëêêùûúú其中ꎬPθt为源项ꎬ其作用是使边界层外部的Re~θt等于Reθtꎬ定义式为Pθt=cθtρt(Reθt-Re~θt)(1.0-Fθt)Reθt采用以下经验公式Reθt=1173.51-589 428Tu+0.2196Tu2æèçöø÷F(λθ)ꎬTuɤ0.3Reθt=331.50(Tu-0.5658)-0.671F(λθ)ꎬTu>0.3ìîíïïïïF(λθ)=1+(12.986λθ+123.66λ2θ+405.689λ3θ)e-(Tu1.5)1.5ꎬ㊀λθɤ0F(λθ)=1+0.275(1-e-35.0λθ)e-(Tu0.5)ꎬλθ>0ìîíïïïï在实际计算中ꎬ通过γ ̄Re~θt转捩模型获得间歇因子ꎬ再通过间歇因子来控制SSTk ̄ω湍流模型中湍动能的生成ꎮγ ̄Re~θt转捩模型与SSTk ̄ω湍流模型耦合为Ə(ρk)Ət+Ə(ρujk)Əxj=γeffτijƏuiƏxj-min(max(γeffꎬ0.1)ꎬ1.0)ρβ∗kω+ƏƏxjμ+μtσkæèçöø÷ƏkƏxjéëêêùûúúƏ(ρω)Ət+Ə(ρujω)Əxj=γvtτijƏuiƏxj-βρω2+ƏƏxj(μ+σωμt)ƏωƏxjéëêêùûúú+2ρ(1-F1)σω21ωƏkƏxjƏωƏxj模型中具体参数定义见文献[23]ꎮ1.3㊀高超声速修正原始SST湍流模型及γ ̄Re~θt转捩模型都是基于不可压缩流动发展的ꎬ为了更好地预测高超声速流动转捩ꎬ本节引入了3种重要的高超声速修正方法ꎮ1.3.1㊀压力梯度修正压力梯度对边界层转捩的影响较大ꎬ在高Mach数情况下ꎬ边界层厚度较大ꎬ进而影响压力梯度的大小ꎬ因此在模拟高超声速流动时应该考虑Mach数对压力梯度的影响ꎮ本文采用张毅峰等[29]提出的压力梯度修正方法ꎬ具体修正形式如下λᶄθ=λθ1+γᶄ-12Maeæèçöø÷其中ꎬMae为边界层外缘Mach数ꎬγᶄ为比热比ꎮ1.3.2㊀高速横流修正在原始γ ̄Re~θt转捩模型中ꎬ没有考虑横流不稳定性对转捩的影响ꎬ对于横流模态主导的转捩ꎬ原始转捩模型计算的结果并不理想ꎮLangtry等[30]在2015年对γ ̄Re~θt转捩模型进行了低速横流修正ꎬ向星皓等[9]在Langtry低速横流修正的基础上ꎬ对高超声速椭圆锥转捩DNS数据进行了拓展ꎬ提出了高速横流转捩判据ꎬ本文直接采用向星皓提出的高速横流转捩方法ꎮLangtry将横流强度引入转捩发生动量厚度Reynolds数输运方程中Ə(ρRe~θt)Ət+Ə(ρujRe~θt)Əxj=Pθt+DSCF+ƏƏxjσθt(μ+μt)ƏRe~θtƏxjéëêêùûúú式中ꎬDSCF为横流源项ꎬLangtry低速横流修正为DSCF=cθtρtccrossflowmin(ReSCF-Re~θtꎬ0.0)Fθt2其中ꎬReSCF为低速横流判据ReSCF=θtρUlocal0.82æèçöø÷μ=-35.088lnhθtæèçöø÷+319.51+f(+ΔHcrossflow)-f(-ΔHcrossflow)其中ꎬh为壁面粗糙度高度ꎬθt为动量厚度ꎬ11气体物理2024年㊀第9卷ΔHcrossflow是横流强度抬升项ꎮ向星皓提出的高速横流转捩判据ꎬ其中高速横流源项DSCF ̄H为DSCF ̄H=cCFρmin(ReSCF ̄H-Re~θtꎬ0)FθtReSCF ̄H=CCF ̄1lnhlμ+CCF ̄2+(Hcrossflow)其中ꎬCCF ̄1=-9.618ꎬCCF ̄2=128.33ꎻlμ为粗糙度参考高度ꎬlμ=1μmꎻf(Hcrossflow)为抬升函数f(Hcrossflow)=60000.1066-ΔHcrossflow+50000(0.1066-ΔHcrossflow)2其中ꎬΔHcrossflow与Langtry低速横流修正中保持一致ꎮ1.3.3㊀SST可压缩修正高超声速流动具有强可压缩性ꎬ所以在进行高超声速计算时ꎬ应该对湍流模型进行可压缩修正ꎮSarkar[31]提出了膨胀耗散修正ꎬ对SST湍流模型中的闭合系数β∗ꎬβ进行了可压缩修正ꎬWilcox[32]在Sarkar修正的基础上考虑了可压缩生成项产生时的延迟效应ꎬ使得可压缩修正在湍流Mach数较小的近壁面关闭ꎬ在湍流Mach数较大的自由剪切层打开ꎬ本文采用Wilcox提出的可压缩性修正β∗=β∗0[1+ξ∗F(Mat)]β=β0-β∗0ξ∗F(Mat)其中ꎬβ0ꎬβ∗均为原始模型中的系数ꎬξ∗=1.5ꎮF(Mat)=[Mat-Mat0]H(Mat-Mat0)Mat0=1/4ꎬH(x)=0ꎬxɤ01ꎬx>0{其中ꎬMat=2k/a为湍流Mach数ꎬa为当地声速ꎮ2㊀网格无关性验证及数值方法确认2.1㊀网格无关性验证计算采用3套网格ꎬ考虑到HyTRV的几何对称性ꎬ生成3套半模网格ꎬ第1层网格高度为1ˑ10-6mꎬ确保y+<1ꎬ流向ˑ法向ˑ周向的网格数分别为:网格1是301ˑ201ˑ201ꎬ网格2是301ˑ301ˑ201ꎬ网格3是401ˑ381ˑ281ꎮ全模下表面如图2所示ꎬ选取y/L=0中心线和x/L=0.5处ꎬ对比3套网格的表面摩阻系数ꎬ计算结果如图3所示ꎮ采用网格1时ꎬ表面摩阻系数分布与另外两个结果存在明显差异ꎻ而采用网格2和网格3时ꎬ表面摩阻系数曲线基本重合ꎬ表明在流向㊁法向和周向均满足网格无关性ꎬ后续数值计算采用网格2ꎮ图2㊀截取位置示意图Fig.2㊀Schematicdiagramoftheinterceptionlocation(a)Surfacefrictionaty/L=0(b)Surfacefrictionatx/L=0.5图3㊀采用3套网格计算得到的摩阻对比Fig.3㊀Comparisonofthefrictiondragcalculatedusingthreesetsofgrids2.2㊀数值方法和自研软件的确认采用修正后的转捩模型对HyTRV开展计算ꎬ计算工况为Ma=6ꎬ来流温度Tɕ=97Kꎬ单位21第2期章录兴ꎬ等:Mach数和壁面温度对HyTRV边界层转捩的影响Reynolds数为Re=1.1ˑ107/mꎬ攻角α=0ʎꎬ来流湍流度FSTI=0.8%ꎬ壁面温度T=300Kꎮ为方便对比分析ꎬ计算结果与参考结果均采用上下对称形式布置ꎬ例如ꎬ图4是模型下表面计算结果与实验结果对比:对于下表面两侧转捩的起始位置ꎬ高超声速修正前的转捩位置在x=0.68m附近ꎬ高超声速修正后的计算结果与实验结果吻合良好ꎬ均在x=0.60m附近ꎬ并且湍流边界层区域形状基本一致ꎬ说明修正后的转捩模型能够较好地预测HyTRV转捩的位置ꎮ(a)Calculationofthefrictiondistribution(beforehypersoniccorrection)(b)Calculationofthefrictiondistribution(afterhypersoniccorrection)(c)Experimentalresultsoftheheatfluxdistribution[17]图4㊀下表面计算结果和实验结果对比Fig.4㊀Comparisonofthecalculatedandexperimentalresultsonthelowersurface3㊀HyTRV转捩的基本流动特性计算工况采用Ma=6ꎬ攻角α=0ʎꎬ来流湍流度FSTI=0.6%ꎬ分析HyTRV转捩的基本流动特性ꎮ从图5可以看出ꎬ模型两侧和顶端均出现高压区ꎬ高压区之间为低压区ꎬ横截面上存在周向压力梯度ꎬ流动从高压区向低压区汇集ꎬ从而在下表面中心线附近和上表面两侧腰部区域均形成流向涡结构(见图6)ꎬ沿流动方向ꎬ高压区域逐渐扩大ꎬ流向涡结构的影响范围也越大ꎮ在流向涡结构的边缘位置ꎬ壁面附近的低速流体被抬升到外壁面区域ꎬ外壁面区域的高速流体又被带入到近壁面区域ꎬ进而导致流向涡结构边缘处壁面的摩阻显著增加ꎬ最终诱发转捩ꎬ这些流动特征与文献[15]的结果一致ꎮ图7显示了上下表面摩阻的分布情况ꎬ其中上表面两侧区域在x/L=0.80附近ꎬ摩阻显著增加ꎬ出现明显的转捩现象ꎬ转捩区域分布在两侧边缘位置ꎻ而下表面两侧区域在x/L=0.75附近ꎬ也出现明显的转捩ꎬ转捩区域相对集中在中心线两侧ꎮ图5㊀不同截面位置处的压力云图Fig.5㊀Pressurecontoursatdifferentcross ̄sectionlocations图6㊀不同截面位置处的流向速度云图Fig.6㊀Streamwisevelocitycontoursatdifferentcross ̄sectionlocations31气体物理2024年㊀第9卷(a)Uppersurface㊀㊀㊀㊀㊀(b)Lowersurface图7㊀上下表面摩阻分布云图Fig.7㊀Frictioncoefficientcontoursontheupperandlowersurfaces4㊀不同Mach数对HyTRV转捩的影响保持来流湍流度FSTI=0.6%不变ꎬMach数变化范围为3~8ꎮ图8是不同Mach数条件下HyTRV上下表面的摩阻分布云图ꎬ从图中可知ꎬ随着Mach数的增加ꎬ上下表面的湍流区域均逐渐减少ꎬ其中上表面两侧转捩起始位置由x/L=0.56附近后移至x/L=0.92附近ꎬ下表面两侧转捩起始位置由x/L=0.48附近后移至x/L=0.99附近ꎬ上下表面两侧转捩起始位置均大幅后移ꎬ说明Mach数对HyTRV转捩的影响很大ꎮuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(a)Ma=3uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(b)Ma=4uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(c)Ma=541第2期章录兴ꎬ等:Mach数和壁面温度对HyTRV边界层转捩的影响uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(d)Ma=6uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(e)Ma=7uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(f)Ma=8图8㊀不同Mach数条件下摩阻系数分布云图Fig.8㊀FrictioncoefficientcontoursatdifferentMachnumbers上表面选取图7中z/L=0.12的位置ꎬ下表面选取z/L=0.10的位置进行分析ꎮ从图9中可以分析出ꎬ随着Mach数的增加ꎬ上表面转捩起始位置不断后移ꎬ当Mach数增加到7时ꎬ由于湍流区的缩小ꎬ此处位置不再发生转捩ꎬ此外ꎬMach数越高层流区摩阻系数越大ꎻ下表面转捩起始位置也不断后移ꎬ当Mach数增加到8时ꎬ此处位置不再发生转捩ꎬ此外ꎬMach数越高ꎬ湍流区的摩阻系数越小ꎬ这些结论与关于来流Mach数对转捩位置影响的普遍研究结论一致ꎮ(a)Uppersurface㊀㊀㊀㊀㊀(b)Lowersurface图9㊀不同位置摩阻系数随Mach数的变化Fig.9㊀VariationoffrictioncoefficientwithMachnumberatdifferentlocations51气体物理2024年㊀第9卷5㊀不同壁面温度对HyTRV转捩的影响保持来流湍流度FSTI=0.6%及Ma=6不变ꎬ壁面温度的变化范围为150~900Kꎮ图10是不同壁面温度条件下HyTRV上下表面的摩阻分布云图ꎬ可以看出随着壁面温度的增加ꎬ上表面两侧湍流区域先是缓慢扩大ꎬ在壁面温度为500K时湍流区域快速缩小ꎬ增加到900K时ꎬ已无明显湍流区域ꎻ下表面两侧湍流区域先是无明显变化ꎬ同样当壁面温度升高到500K时ꎬ湍流区域快速缩小ꎬ当壁面温度升高到700K时ꎬ两侧已经无明显的湍流区域ꎬ相比上表面两侧湍流区域ꎬ下表面湍流区域消失得更早ꎮ由此可以得出壁面温度对转捩的产生有较大的影响ꎬ壁面温度增加到一定程度将导致HyTRV没有明显的转捩现象ꎮuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(a)T=150Kuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(b)T=200Kuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(c)T=300Kuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(d)T=500K61第2期章录兴ꎬ等:Mach数和壁面温度对HyTRV边界层转捩的影响uppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(e)T=700Kuppersurface㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀lowersurface(f)T=900K图10㊀不同壁面温度条件下摩阻系数分布云图Fig.10㊀Frictioncoefficientcontoursatdifferentwalltemperatureconditions上表面选取z/L=0.125的位置ꎬ下表面选取z/L=0.100的位置进行分析ꎮ从图11中可以分析出ꎬ随着壁面温度的增加ꎬ上表面转捩起始位置先前移ꎬ当壁面温度增加到500K时ꎬ转捩起始位置后移ꎬ转捩区长度逐渐增加ꎬ层流区域的摩阻系数逐渐增加ꎬ当壁面温度增加到700K时ꎬ该位置已不再出现转捩ꎻ下表面转捩起始位置先小幅后移ꎬ当壁面温度增加到300K时ꎬ转捩起始位置开始后移ꎬ当壁面温度增加到700K时ꎬ由于湍流区域的减小ꎬ该位置不再发生转捩ꎮ(a)Uppersurface㊀㊀㊀㊀㊀(b)Lowersurface图11㊀不同位置摩阻系数随壁面温度的变化Fig.11㊀Variationoffrictioncoefficientwithwalltemperatureatdifferentlocations为进一步分析壁面温度的影响ꎬ本文分别在上下表面湍流区选取一点(0.9ꎬ0.029ꎬ0.14)ꎬ(0.97ꎬ-0.34ꎬ0.12)ꎬ分析边界层湍动能剖面ꎬ结果如图12所示ꎮ从图中可以看到ꎬ随着壁面温度升高ꎬ边界层厚度先略微变厚ꎬ再变薄ꎬ当壁面温度升高到700K时ꎬ边界层厚度迅速降低ꎮ这些结果与转捩位置先前移再后移的结论相符合ꎬ因为边界层厚度会影响不稳定波的时间和空间尺度ꎬ边界层厚度低时ꎬ不稳定波增长速度变慢ꎬ延迟转捩发生ꎮ需要指出的是ꎬ仅采用当前使用的方法ꎬ无法从更深层71气体物理2024年㊀第9卷次揭示转捩反转的流动机理ꎬ而须另外借助稳定性分析方法ꎬ例如ꎬ使用eN方法开展基于模态的稳定性研究ꎮ文献[33]采用该手段研究了大掠角平板钝三角翼随壁温比变化出现转捩反转的内在机理:壁温比升高促进横流模态和第1模态扰动增长ꎬ抑制第2模态发展ꎬ在第1㊁2模态联合作用影响下ꎬ出现转捩反转现象ꎮ我们将在后续开展进一步研究ꎮ(a)Uppersurface(b)Lowersurface图12㊀不同位置湍动能剖面随壁面温度的变化Fig.12㊀Variationofturbulentkineticenergywithwalltemperatureatdifferentlocations6㊀结论针对HyTRV转捩问题ꎬ在Mach数Ma=3~8ꎬ壁面温度T=150~900K的条件下ꎬ基于课题组自研软件ꎬ对γ ̄Re~θt转捩模型和SST湍流模型进行了高超声速修正ꎬ研究了Mach数和壁面温度对HyTRV转捩的影响ꎬ得出以下结论:1)经过高超声速修正后的γ ̄Re~θt转捩模型和SST湍流模型能够较为准确地预测HyTRV转捩位置ꎬ并且湍流边界层区域形状与实验结果基本一致ꎻHyTRV存在多个不同的转捩区域ꎬ上表面两侧转捩区域分布在两侧边缘位置ꎬ下表面两侧转捩区域分布在中心线两侧ꎮ2)Mach数的增加会导致上下表面转捩起始位置均大幅后移ꎬ湍流区大幅缩小ꎬ但当Mach数增加到8时ꎬ湍流区仍然存在ꎬ并没有消失ꎻ上表面层流区摩阻不断增加ꎬ下表面湍流区摩阻不断减小ꎮ3)壁面温度的增加会导致上下表面转捩起始位置先前移ꎬ再后移ꎬ这与边界层厚度变化规律一致ꎬ当壁面温度增加到700K时ꎬ下表面湍流区已经基本消失ꎬ当壁面温度增加到900K时ꎬ上表面湍流区也基本消失ꎻ上表面在层流区域的摩阻系数逐渐增大ꎬ在湍流区的摩阻系数逐渐减小ꎮ致谢㊀感谢中国空气动力研究与发展中心和空天飞行空气动力科学与技术全国重点实验室提供的HyTRV模型数据和实验数据ꎮ参考文献(References)[1]㊀OberingIIIHꎬHeinrichsRL.Missiledefenseforgreatpowerconflict:outmaneuveringtheChinathreat[J].Stra ̄tegicStudiesQuarterlyꎬ2019ꎬ3(4):37 ̄56. [2]ChengCꎬWuJHꎬZhangYLꎬetal.Aerodynamicsanddynamicstabilityofmicro ̄air ̄vehiclewithfourflappingwingsinhoveringflight[J].AdvancesinAerodynamicsꎬ2020ꎬ2(3):5.[3]BertinJJꎬCummingsRM.Fiftyyearsofhypersonics:whereweᶄvebeenꎬwhereweᶄregoing[J].ProgressinAerospaceSciencesꎬ2003ꎬ39(6/7):511 ̄536. [4]陈坚强ꎬ涂国华ꎬ张毅锋ꎬ等.高超声速边界层转捩研究现状与发展趋势[J].空气动力学学报ꎬ2017ꎬ35(3):311 ̄337.ChenJQꎬTuGHꎬZhangYFꎬetal.Hypersnonicboundarylayertransition:whatweknowꎬwhereshallwego[J].ActaAerodynamicaSinicaꎬ2017ꎬ35(3):311 ̄337(inChinese).[5]段毅ꎬ姚世勇ꎬ李思怡ꎬ等.高超声速边界层转捩的若干问题及工程应用研究进展综述[J].空气动力学学报ꎬ2020ꎬ38(2):391 ̄403.DuanYꎬYaoSYꎬLiSYꎬetal.Reviewofprogressinsomeissuesandengineeringapplicationofhypersonicboundarylayertransition[J].ActaAerodynamicaSinicaꎬ2020ꎬ38(2):391 ̄403(inChinese).[6]ZhangYFꎬZhangYRꎬChenJQꎬetal.Numericalsi ̄81第2期章录兴ꎬ等:Mach数和壁面温度对HyTRV边界层转捩的影响mulationsofhypersonicboundarylayertransitionbasedontheflowsolverchant2.0[R].AIAA2017 ̄2409ꎬ2017. [7]KrauseMꎬBehrMꎬBallmannJ.Modelingoftransitioneffectsinhypersonicintakeflowsusingacorrelation ̄basedintermittencymodel[R].AIAA2008 ̄2598ꎬ2008. [8]YiMRꎬZhaoHYꎬLeJL.Hypersonicnaturalandforcedtransitionsimulationbycorrelation ̄basedintermit ̄tency[R].AIAA2017 ̄2337ꎬ2017.[9]向星皓ꎬ张毅锋ꎬ袁先旭ꎬ等.C ̄γ ̄Reθ高超声速三维边界层转捩预测模型[J].航空学报ꎬ2021ꎬ42(9):625711.XiangXHꎬZhangYFꎬYuanXXꎬetal.C ̄γ ̄Reθmodelforhypersonicthree ̄dimensionalboundarylayertransitionprediction[J].ActaAeronauticaetAstronauticaSinicaꎬ2021ꎬ42(9):625711(inChinese).[10]McDanielRDꎬNanceRPꎬHassanHA.Transitiononsetpredictionforhigh ̄speedflow[J].JournalofSpacecraftandRocketsꎬ2000ꎬ37(3):304 ̄309.[11]PappJLꎬDashSM.Rapidengineeringapproachtomodelinghypersoniclaminar ̄to ̄turbulenttransitionalflows[J].JournalofSpacecraftandRocketsꎬ2005ꎬ42(3):467 ̄475.[12]JulianoTJꎬSchneiderSP.InstabilityandtransitionontheHIFiRE ̄5inaMach ̄6quiettunnel[R].AIAA2010 ̄5004ꎬ2010.[13]杨云军ꎬ马汉东ꎬ周伟江.高超声速流动转捩的数值研究[J].宇航学报ꎬ2006ꎬ27(1):85 ̄88.YangYJꎬMaHDꎬZhouWJ.Numericalresearchonsupersonicflowtransition[J].JournalofAstronauticsꎬ2006ꎬ27(1):85 ̄88(inChinese).[14]袁先旭ꎬ何琨ꎬ陈坚强ꎬ等.MF ̄1模型飞行试验转捩结果初步分析[J].空气动力学学报ꎬ2018ꎬ36(2):286 ̄293.YuanXXꎬHeKꎬChenJQꎬetal.PreliminarytransitionresearchanalysisofMF ̄1[J].ActaAerodynamicaSinicaꎬ2018ꎬ36(2):286 ̄293(inChinese).[15]陈坚强ꎬ涂国华ꎬ万兵兵ꎬ等.HyTRV流场特征与边界层稳定性特征分析[J].航空学报ꎬ2021ꎬ42(6):124317.ChenJQꎬTuGHꎬWanBBꎬetal.Characteristicsofflowfieldandboundary ̄layerstabilityofHyTRV[J].ActaAeronauticaetAstronauticaSinicaꎬ2021ꎬ42(6):124317(inChinese).[16]陈坚强ꎬ刘深深ꎬ刘智勇ꎬ等.用于高超声速边界层转捩研究的标模气动布局及设计方法.中国:109969374B[P].2021 ̄05 ̄18.ChenJQꎬLiuSSꎬLiuZYꎬetal.Standardmodelaero ̄dynamiclayoutanddesignmethodforhypersonicboundarylayertransitionresearch.CNꎬ109969374B[P].2021 ̄05 ̄18(inChinese).[17]LiuSSꎬYuanXXꎬLiuZYꎬetal.Designandtransitioncharacteristicsofastandardmodelforhypersonicboundarylayertransitionresearch[J].ActaMechanicaSinicaꎬ2021ꎬ37(11):1637 ̄1647.[18]ChenXꎬDongSWꎬTuGHꎬetal.Boundarylayertran ̄sitionandlinearmodalinstabilitiesofhypersonicflowoveraliftingbody[J].JournalofFluidMechanicsꎬ2022ꎬ938(408):A8.[19]QiHꎬLiXLꎬYuCPꎬetal.Directnumericalsimulationofhypersonicboundarylayertransitionoveralifting ̄bodymodelHyTRV[J].AdvancesinAerodynamicsꎬ2021ꎬ3(1):31.[20]万兵兵ꎬ陈曦ꎬ陈坚强ꎬ等.三维边界层转捩预测HyTEN软件在高超声速典型标模中的应用[J].空天技术ꎬ2023(1):150 ̄158.WanBBꎬChenXꎬChenJQꎬetal.ApplicationsofHyTENsoftwareforpredictingthree ̄dimensionalboundary ̄layertransitionintypicalhypersonicmodels[J].AerospaceTechnologyꎬ2023(1):150 ̄158(inChinese). [21]MenterFRꎬLangtryRBꎬLikkiSRꎬetal.Acorrelation ̄basedtransitionmodelusinglocalvariables PartⅠ:modelformulation[J].JournalofTurbomachineryꎬ2006ꎬ128(3):413 ̄422.[22]LangtryRBꎬMenterFRꎬLikkiSRꎬetal.Acorrelation ̄basedtransitionmodelusinglocalvariables PartⅡ:testcasesandindustrialapplications[J].JournalofTur ̄bomachineryꎬ2006ꎬ128(3):423 ̄434.[23]LangtryRBꎬMenterFR.Correlation ̄basedtransitionmodelingforunstructuredparallelizedcomputationalfluiddynamicscodes[J].AIAAJournalꎬ2009ꎬ47(12):2894 ̄2906.[24]孟德虹ꎬ张玉伦ꎬ王光学ꎬ等.γ ̄Reθ转捩模型在二维低速问题中的应用[J].航空学报ꎬ2011ꎬ32(5):792 ̄801.MengDHꎬZhangYLꎬWangGXꎬetal.Applicationofγ ̄Reθtransitionmodeltotwo ̄dimensionallowspeedflows[J].ActaAeronauticaetAstronauticaSinicaꎬ2011ꎬ32(5):792 ̄801(inChinese).[25]牟斌ꎬ江雄ꎬ肖中云ꎬ等.γ ̄Reθ转捩模型的标定与应用[J].空气动力学学报ꎬ2013ꎬ31(1):103 ̄109.MouBꎬJiangXꎬXiaoZYꎬetal.Implementationandcaliberationofγ ̄Reθtransitionmodel[J].ActaAerody ̄namicaSinicaꎬ2013ꎬ31(1):103 ̄109(inChinese). [26]郭隽ꎬ刘丽平ꎬ徐晶磊ꎬ等.γ ̄Re~θt转捩模型在跨声速涡轮叶栅中的应用[J].推进技术ꎬ2018ꎬ39(9):1994 ̄2001.91气体物理2024年㊀第9卷GuoJꎬLiuLPꎬXuJLꎬetal.Applicationofγ ̄Re~θttran ̄sitionmodelintransonicturbinecascades[J].JournalofPropulsionTechnologyꎬ2018ꎬ39(9):1994 ̄2001(inChinese).[27]郑赟ꎬ李虹杨ꎬ刘大响.γ ̄Reθ转捩模型在高超声速下的应用及分析[J].推进技术ꎬ2014ꎬ35(3):296 ̄304.ZhengYꎬLiHYꎬLiuDX.Applicationandanalysisofγ ̄Reθtransitionmodelinhypersonicflow[J].JournalofPropulsionTechnologyꎬ2014ꎬ35(3):296 ̄304(inChi ̄nese).[28]孔维萱ꎬ阎超ꎬ赵瑞.γ ̄Reθ模式应用于高速边界层转捩的研究[J].空气动力学学报ꎬ2013ꎬ31(1):120 ̄126.KongWXꎬYanCꎬZhaoR.γ ̄Reθmodelresearchforhigh ̄speedboundarylayertransition[J].ActaAerody ̄namicaSinicaꎬ2013ꎬ31(1):120 ̄126(inChinese). [29]张毅锋ꎬ何琨ꎬ张益荣ꎬ等.Menter转捩模型在高超声速流动模拟中的改进及验证[J].宇航学报ꎬ2016ꎬ37(4):397 ̄402.ZhangYFꎬHeKꎬZhangYRꎬetal.ImprovementandvalidationofMenterᶄstransitionmodelforhypersonicflowsimulation[J].JournalofAstronauticsꎬ2016ꎬ37(4):397 ̄402(inChinese).[30]LangtryRBꎬSenguptaKꎬYehDTꎬetal.Extendingtheγ ̄Reθtcorrelationbasedtransitionmodelforcrossfloweffects(Invited)[R].AIAA2015 ̄2474ꎬ2015. [31]SarkarS.Thepressure ̄dilatationcorrelationincompressi ̄bleflows[J].PhysicsofFluidsAꎬ1992ꎬ4(12):2674 ̄2682.[32]WilcoxDC.Dilatation ̄dissipationcorrectionsforadvancedturbulencemodels[J].AIAAJournalꎬ1992ꎬ30(11):2639 ̄2646.[33]马祎蕾ꎬ余平ꎬ姚世勇.壁温对钝三角翼边界层稳定性及转捩影响[J].空气动力学学报ꎬ2020ꎬ38(6):1017 ̄1026.MaYLꎬYuPꎬYaoSY.Effectofwalltemperatureonstabilityandtransitionofhypersonicboundarylayeronabluntdeltawing[J].ActaAerodynamicaSinicaꎬ2020ꎬ38(6):1017 ̄1026(inChinese).02。
计算机考研--复试英语Abstract 1In recent years, machine learning has developed rapidly, especially in the deep learning, where remarkable achievements are obtained in image, voice, natural language processing and other fields. The expressive ability of machine learning algorithm has been greatly improved; however, with the increase of model complexity, the interpretability of computer learning algorithm has deteriorated. So far, the interpretability of machine learning remains as a challenge. The trained models via algorithms are regarded as black boxes, which seriously hamper the use of machine learning in certain fields, such as medicine, finance and so on. Presently, only a few works emphasis on the interpretability of machine learning. Therefore, this paper aims to classify, analyze and compare the existing interpretable methods; on the one hand, it expounds the definition and measurement of interpretability, while on the other hand, for the different interpretable objects, it summarizes and analyses various interpretable techniques of machine learning from three aspects: model understanding, prediction result interpretation and mimic model understanding. Moreover, the paper also discusses the challenges and opportunities faced by machine learning interpretable methods and the possible development direction in the future. The proposed interpretation methods should also be useful for putting many research open questions in perspective.摘要近年来,机器学习发展迅速,尤其是深度学习在图像、声⾳、⾃然语⾔处理等领域取得卓越成效.机器学习算法的表⽰能⼒⼤幅度提⾼,但是伴随着模型复杂度的增加,机器学习算法的可解释性越差,⾄今,机器学习的可解释性依旧是个难题.通过算法训练出的模型被看作成⿊盒⼦,严重阻碍了机器学习在某些特定领域的使⽤,譬如医学、⾦融等领域.⽬前针对机器学习的可解释性综述性的⼯作极少,因此,将现有的可解释⽅法进⾏归类描述和分析⽐较,⼀⽅⾯对可解释性的定义、度量进⾏阐述,另⼀⽅⾯针对可解释对象的不同,从模型的解释、预测结果的解释和模仿者模型的解释3个⽅⾯,总结和分析各种机器学习可解释技术,并讨论了机器学习可解释⽅法⾯临的挑战和机遇以及未来的可能发展⽅向.Abstract 2Deep learning is an important field of machine learning research, which is widely used in industry for its powerful feature extraction capabilities and advanced performance in many applications. However, due to the bias in training data labeling and model design, research shows that deep learning may aggravate human bias and discrimination in some applications, which results in unfairness during the decision-making process, thereby will cause negative impact to both individuals and socials. To improve the reliability of deep learning and promote its development in the field of fairness, we review the sources of bias in deep learning, debiasing methods for different types biases, fairness measure metrics for measuring the effect of debiasing, and current popular debiasing platforms, based on the existing research work. In the end we explore the open issues in existing fairness research field and future development trends.摘要:深度学习是机器学习研究中的⼀个重要领域,它具有强⼤的特征提取能⼒,且在许多应⽤中表现出先进的性能,因此在⼯业界中被⼴泛应⽤.然⽽,由于训练数据标注和模型设计存在偏见,现有的研究表明深度学习在某些应⽤中可能会强化⼈类的偏见和歧视,导致决策过程中的不公平现象产⽣,从⽽对个⼈和社会产⽣潜在的负⾯影响.为提⾼深度学习的应⽤可靠性、推动其在公平领域的发展,针对已有的研究⼯作,从数据和模型2⽅⾯出发,综述了深度学习应⽤中的偏见来源、针对不同类型偏见的去偏⽅法、评估去偏效果的公平性评价指标、以及⽬前主流的去偏平台,最后总结现有公平性研究领域存在的开放问题以及未来的发展趋势.Abstract 3TensorFlow Lite (TFLite) is a lightweight, fast and cross-platform open source machine learning framework specifically designed for mobile and IoT. It’s part of TensorFlow and supports multiple platforms such as Android, iOS, embedded Linux, and MCU etc. It greatly reduces the barrier for developers, accelerates the development of on-device machine learning (ODML), and makes ML run everywhere. This article introduces the trend, challenges and typical applications of ODML; the origin and system architecture of TFLite; best practices and tool chains suitable for ML beginners; and the roadmap of TFLite.摘要: TensorFlow Lite(TFLite)是⼀个轻量、快速、跨平台的专门针对移动和IoT场景的开源机器学习框架,是TensorFlow 的⼀部分,⽀持安卓、iOS、嵌⼊式Linux以及MCU等多个平台部署.它⼤⼤降低开发者使⽤门槛,加速端侧机器学习的发展,推动机器学习⽆处不在.介绍了端侧机器学习的浪潮、挑战和典型应⽤;TFLite的起源和系统架构;TFLite的最佳实践,以及适合初学者的⼯具链;展望了未来的发展⽅向.Abstract 4The rapid development of the Internet accesses many new applications including real time multi-media service, remote cloud service, etc. These applications require various types of service quality, which is a significant challenge towards current best effort routing algorithms. Since the recent huge success in applying machine learning in game, computervision and natural language processing, many people tries to design “smart” routing algorithms based on machine learning methods. In contrary with traditional model-based, decentralized routing algorithms (e.g.OSPF), machine learning based routing algorithms are usually data-driven, which can adapt to dynamically changing network environments and accommodate different service quality requirements. Data-driven routing algorithms based on machine learning approach have shown great potential in becoming an important part of the next generation network. However, researches on artificial intelligent routing are still on a very beginning stage. In this paper we firstly introduce current researches on data-driven routing algorithms based on machine learning approach, showing the main ideas, application scenarios and pros and cons of these different works. Our analysis shows that current researches are mainly for the principle of machine learning based routing algorithms but still far from deployment in real scenarios. So we then analyze different training and deploying methods for machine learning based routing algorithms in real scenarios and propose two reasonable approaches to train and deploy such routing algorithms with low overhead and high reliability. Finally, we discuss the opportunities and challenges and show several potential research directions for machine learning based routing algorithms in the future.摘要:互联⽹的飞速发展催⽣了很多新型⽹络应⽤,其中包括实时多媒体流服务、远程云服务等.现有尽⼒⽽为的路由转发算法难以满⾜这些应⽤所带来的多样化的⽹络服务质量需求.随着近些年将机器学习⽅法应⽤于游戏、计算机视觉、⾃然语⾔处理获得了巨⼤的成功,很多⼈尝试基于机器学习⽅法去设计智能路由算法.相⽐于传统数学模型驱动的分布式路由算法⽽⾔,基于机器学习的路由算法通常是数据驱动的,这使得其能够适应动态变化的⽹络环境以及多样的性能评价指标优化需求.基于机器学习的数据驱动智能路由算法⽬前已经展⽰出了巨⼤的潜⼒,未来很有希望成为下⼀代互联⽹的重要组成部分.然⽽现有对于智能路由的研究仍然处于初步阶段.⾸先介绍了现有数据驱动智能路由算法的相关研究,展现了这些⽅法的核⼼思想和应⽤场景并分析了这些⼯作的优势与不⾜.分析表明,现有基于机器学习的智能路由算法研究主要针对算法原理,这些路由算法距离真实环境下部署仍然很遥远.因此接下来分析了不同的真实场景智能路由算法训练和部署⽅案并提出了2种合理的训练部署框架以使得智能路由算法能够低成本、⾼可靠性地在真实场景被部署.最后分析了基于机器学习的智能路由算法未来发展中所⾯临的机遇与挑战并给出了未来的研究⽅向.Abstract 5In recent years, the rapid development of Internet technology has greatly facilitated the daily life of human, and it is inevitable that massive information erupts in a blowout. How to quickly and effectively obtain the required information on the Internet is an urgent problem. The automatic text summarization technology can effectively alleviate this problem. As one of the most important fields in natural language processing and artificial intelligence, it can automatically produce a concise and coherent summary from a long text or text set through computer, in which the summary should accurately reflect the central themes of source text. In this paper, we expound the connotation of automatic summarization, review the development of automatic text summarization technique and introduce two main techniques in detail: extractive and abstractive summarization, including feature scoring, classification method, linear programming, submodular function, graph ranking, sequence labeling, heuristic algorithm, deep learning, etc. We also analyze the datasets and evaluation metrics that are commonly used in automatic summarization. Finally, the challenges ahead and the future trends of research and application have been predicted.摘要:近年来,互联⽹技术的蓬勃发展极⼤地便利了⼈类的⽇常⽣活,不可避免的是互联⽹中的信息呈井喷式爆发,如何从中快速有效地获取所需信息显得极为重要.⾃动⽂本摘要技术的出现可以有效缓解该问题,其作为⾃然语⾔处理和⼈⼯智能领域的重要研究内容之⼀,利⽤计算机⾃动地从长⽂本或⽂本集合中提炼出⼀段能准确反映源⽂中⼼内容的简洁连贯的短⽂.探讨⾃动⽂本摘要任务的内涵,回顾和分析了⾃动⽂本摘要技术的发展,针对⽬前主要的2种摘要产⽣形式(抽取式和⽣成式)的具体⼯作进⾏了详细介绍,包括特征评分、分类算法、线性规划、次模函数、图排序、序列标注、启发式算法、深度学习等算法.并对⾃动⽂本摘要常⽤的数据集以及评价指标进⾏了分析,最后对其⾯临的挑战和未来的研究趋势、应⽤等进⾏了预测.Abstract 6With the high-speed development of Internet of things, wearable devices and mobile communication technology, large-scale data continuously generate and converge to multiple data collectors, which influences people’s life in many ways. Meanwhile, it also causes more and more severe privacy leaks. Traditional privacy aware mechanisms such as differential privacy, encryption and anonymization are not enough to deal with the serious situation. What is more, the data convergence leads to data monopoly which hinders the realization of the big data value seriously. Besides, tampered data, single point failure in data quality management and so on may cause untrustworthy data-driven decision-making. How to use big data correctly has become an important issue. For those reasons, we propose the data transparency, aiming to provide solution for the correct use of big data. Blockchain originated from digital currency has the characteristics of decentralization, transparency and immutability, and it provides an accountable and secure solution for data transparency. In this paper, we first propose the definition and research dimension of the data transparency from the perspective of big data life cycle, and we also analyze and summary the methods to realize data transparency. Then, we summary the research progress of blockchain-based data transparency. Finally, we analyze the challenges that may arise in the process of blockchain-based data transparency.摘要:物联⽹、穿戴设备和移动通信等技术的⾼速发展促使数据源源不断地产⽣并汇聚⾄多⽅数据收集者,由此带来更严峻的隐私泄露问题, 然⽽传统的差分隐私、加密和匿名等隐私保护技术还不⾜以应对.更进⼀步,数据的⾃主汇聚导致数据垄断问题,严重影响了⼤数据价值实现.此外,⼤数据决策过程中,数据⾮真实产⽣、被篡改和质量管理过程中的单点失败等问题导致数据决策不可信.如何使这些问题得到有效治理,使数据被正确和规范地使⽤是⼤数据发展⾯临的主要挑战.⾸先,提出数据透明化的概念和研究框架,旨在增加⼤数据价值实现过程的透明性,从⽽为上述问题提供解决⽅案.然后,指出数据透明化的实现需求与区块链的特性天然契合,并对⽬前基于区块链的数据透明化研究现状进⾏总结.最后,对基于区块链的数据透明化可能⾯临的挑战进⾏分析.Abstract 7Blockchain technology is a new emerging technology that has the potential to revolutionize many traditional industries. Since the creation of Bitcoin, which represents blockchain 1.0, blockchain technology has been attracting extensive attention and a great amount of user transaction data has been accumulated. Furthermore, the birth of Ethereum, which represents blockchain 2.0, further enriches data type in blockchain. While the popularity of blockchain technology bringing about a lot of technical innovation, it also leads to many new problems, such as user privacy disclosure and illegal financial activities. However, the public accessible of blockchain data provides unprecedented opportunity for researchers to understand and resolve these problems through blockchain data analysis. Thus, it is of great significance to summarize the existing research problems, the results obtained, the possible research trends, and the challenges faced in blockchain data analysis. To this end, a comprehensive review and summary of the progress of blockchain data analysis is presented. The review begins by introducing the architecture and key techniques of blockchain technology and providing the main data types in blockchain with the corresponding analysis methods. Then, the current research progress in blockchain data analysis is summarized in seven research problems, which includes entity recognition, privacy disclosure risk analysis, network portrait, network visualization, market effect analysis, transaction pattern recognition, illegal behavior detection and analysis. Finally, the directions, prospects and challenges for future research are explored based on the shortcomings of current research.摘要:区块链是⼀项具有颠覆许多传统⾏业的潜⼒的新兴技术.⾃以⽐特币为代表的区块链1.0诞⽣以来,区块链技术获得了⼴泛的关注,积累了⼤量的⽤户交易数据.⽽以以太坊为代表的区块链2.0的诞⽣,更加丰富了区块链的数据类型.区块链技术的⽕热,催⽣了⼤量基于区块链的技术创新的同时也带来许多新的问题,如⽤户隐私泄露,⾮法⾦融活动等.⽽区块链数据公开的特性,为研究⼈员通过分析区块链数据了解和解决相关问题提供了前所未有的机会.因此,总结⽬前区块链数据存在的研究问题、取得的分析成果、可能的研究趋势以及⾯临的挑战具有重要意义.为此,全⾯回顾和总结了当前的区块链数据分析的成果,在介绍区块链技术架构和关键技术的基础上,分析了⽬前区块链系统中主要的数据类型,总结了⽬前区块链数据的分析⽅法,并就实体识别、隐私泄露风险分析、⽹络画像、⽹络可视化、市场效应分析、交易模式识别、⾮法⾏为检测与分析等7个问题总结了当前区块链数据分析的研究进展.最后针对⽬前区块链数据分析研究中存在的不⾜分析和展望了未来的研究⽅向以及⾯临的挑战.Abstract 8In recent years, as more and more large-scale scientific facilities have been built and significant scientific experiments have been carried out, scientific research has entered an unprecedented big data era. Scientific research in big data era is a process of big science, big demand, big data, big computing, and big discovery. It is of important significance to develop a full life cycle data management system for scientific big data. In this paper, we first introduce the background of the development of scientific big data management system. Then we specify the concepts and three key characteristics of scientific big data. After an review of scientific data resource development projects and scientific data management systems, a framework is proposed aiming at the full life cycle management of scientific big data. Further, we introduce the key technologies of the management framework including data fusion, real-time analysis, long termstorage, cloud service, and data opening and sharing. Finally, we summarize the research progress in this field, and look into the application prospects of scientific big data management system.摘要:近年来,随着越来越多的⼤科学装置的建设和重⼤科学实验的开展,科学研究进⼊到⼀个前所未有的⼤数据时代.⼤数据时代科学研究是⼀个⼤科学、⼤需求、⼤数据、⼤计算、⼤发现的过程,研发⼀个⽀持科学⼤数据全⽣命周期的数据管理系统具有重要的意义.分析了研发科学⼤数据管理系统的背景,阐述了科学⼤数据的概念和三⼤特征,通过对科学数据资源发展和科学数据管理系统的研究进展进⾏综述分析,提出了满⾜科学数据管理全⽣命周期的科学⼤数据管理框架,并从数据融合、数据实时分析、长期存储、云服务体系以及数据开放共享机制5个⽅⾯分析了科学⼤数据管理系统中的关键技术.最后,结合科学研究领域展望了科学⼤数据管理系统的应⽤前景.Abstract 9Recently, research on deep learning applied to cyberspace security has caused increasing academic concern, and this survey analyzes the current research situation and trends of deep learning applied to cyberspace security in terms of classification algorithms, feature extraction and learning performance. Currently deep learning is mainly applied to malware detection and intrusion detection, and this survey reveals the existing problems of these applications: feature selection,which could be achieved by extracting features from raw data; self-adaptability, achieved by early-exit strategy to update the model in real time; interpretability, achieved by influence functions to obtain the correspondence between features and classification labels. Then, top 10 obstacles and opportunities in deep learning research are summarized. Based on this, top 10 obstacles and opportunities of deep learning applied to cyberspace security are at first proposed, which falls into three categories. The first category is intrinsic vulnerabilities of deep learning to adversarial attacks and privacy-theft attacks. The second category is sequence-model related problems, including program syntax analysis, program code generation and long-term dependences in sequence modeling. The third category is learning performance problems, including poor interpretability and traceability, poor self-adaptability and self-learning ability, false positives and data unbalance. Main obstacles and their opportunities among the top 10 are analyzed, and we also point out that applications using classification models are vulnerable to adversarial attacks and the most effective solution is adversarial training; collaborative deep learning applications are vulnerable to privacy-theft attacks, and prospective defense is teacher-student model. Finally, future research trends of deep learning applied to cyberspace security are introduced.摘要:近年来,深度学习应⽤于⽹络空间安全的研究逐渐受到国内外学者的关注,从分类算法、特征提取和学习效果等⽅⾯分析了深度学习应⽤于⽹络空间安全领域的研究现状与进展.⽬前,深度学习主要应⽤于恶意软件检测和⼊侵检测两⼤⽅⾯,指出了这些应⽤存在的问题:特征选择问题,需从原始数据中提取更全⾯的特征;⾃适应性问题,可通过early-exit策略对模型进⾏实时更新;可解释性问题,可使⽤影响函数得到特征与分类标签之间的相关性.其次,归纳总结了深度学习发展⾯临的⼗⼤问题与机遇,在此基础上,⾸次归纳了深度学习应⽤于⽹络空间安全所⾯临的⼗⼤问题与机遇,并将⼗⼤问题与机遇归为3类:1)算法脆弱性问题,包括深度学习模型易受对抗攻击和隐私窃取攻击;2)序列化模型相关问题,包括程序语法分析、程序代码⽣成和序列建模长期依赖问题;3)算法性能问题,即可解释性和可追溯性问题、⾃适应性和⾃学习性问题、存在误报以及数据集不均衡的问题.对⼗⼤问题与机遇中主要问题及其解决⽅案进⾏了分析,指出对于分类的应⽤易受对抗攻击,最有效的防御⽅案是对抗训练;基于协作性深度学习进⾏分类的安全应⽤易受隐私窃取攻击,防御的研究⽅向是教师学⽣模型.最后,指出了深度学习应⽤于⽹络空间安全未来的研究发展趋势.。
第35卷第2期2021年2月中文信息学报JO U R N A L OF CHINESE INFO R M A TIO N PROCESSINGVol. 35,No. 2 Feb.,2021文章编号:1003-0077(2021)02-0099-08U -N e t :用于包含无答案问题的机器阅读理解的轻量级模型孙付1,李林阳1,邱锡鹏1,刘扬黄萱菁1(1.复旦大学计算机学院,上海201210; 2.流利说硅谷人工智能实验室,美国旧金山94104)摘要:处理机器阅读理解任务时.识别其中没有答案的问题是自然语言处理领域的一个新的挑战。
该文提出U -N e t 模型来处理这个问题,该模型包括3个主要成分:答案预测模块、无答案判别模块和答案验证模块。
该模型用一个U 节点将问题和文章拼接为一个连续的文本序列,该U 节点同时编码问题和文章的信息,在判断问题是否 有答案时起到重要作用,同时对于精简U -N e t 的结构也有重要作用。
与基于预训练的B E R T 不同,U -N e t 的U 节 点的信息获取方式更多样,并且不需要巨大的计算资源就能有效地完成机器阅读理解任务。
在SQuAD 2.0中,11-他1的单模型['1得分72.6、£1^得分69.3,11-仏1的集成模型/;'|得分74.9、£\1得分71.4.均为公开的非基于大规模预训练语言模型的模型结果的第一名。
关键词:机器阅读理解;SQ uAD ;注意力机制 中图分类号:TP391文献标识码:AU-Net : Lightweight Model for Machine Reading Comprehensionwith Unanswerable QuestionsSUN Fu1 . LI Linyang1 . QIU Xipeng1 , LIU Yang2 , HUANG Xuanjing1(1. School of Computer Science, Fudan University,Shanghai 201210,China;2. Liulishuo Silicon Valley AI Lab, San Francisco 94104» USA)Abstract : Machine reading comprehension with unanswerable questions is a challenging task. In this paper, we propose a unified models called U-Net» with three important components : answer pointer^ no-answer pointer, and answer verifier. We introduce a universal node which processes the question and its context passage as a single contigu- ous sequence of tokens. The universal node encodes the fused information from both the quev s tion and passage» and plays an important role to predict whether the question is answerable and also greatly improves the conciseness of the U-Net. Different from the modeLs based on pre-trained B E R T , universal node fuses information from passage and question in a variety of ways and avoids the huge computation. T he single U-Net model achieves the F\ score of 72.6 and the EM score of 69.3 on SQuAD 2.0, and the ensemble version, 74.9 and 71.4» respectively. Both version of U-Net models rank top among the models without a large scale pre-trained language model.Keywords : machine reading comprehension;SQuAD;attention mechanism了丰硕的成果。
HIDDEN-ARTICULATOR MARKOV MODELS: PERFORMANCE IMPROVEMENTS AND ROBUSTNESS TO NOISE
Matt Richardson, Jeff Bilmes and Chris Diorio University of Washington {mattr@cs, bilmes@ee, diorio@cs}.washington.edu
ABSTRACT A Hidden-Articulator Markov Model (HAMM) is a Hidden Markov Model (HMM) in which each state represents an articu-latory configuration. Articulatory knowledge, known to be use-ful for speech recognition [4], is represented by specifying a mapping of phonemes to articulatory configurations; vocal tract dynamics are represented via transitions between articulatory configurations.
In previous work [13], we extended the articulatory-feature model introduced by Erler [7] by using diphone units and a new technique for model initialization. By comparing it with a purely random model, we showed that the HAMM can take advantage of articulatory knowledge.
In this paper, we extend that work in three ways. First, we de-crease the number of parameters, making it comparable in size to standard HMMs. Second, we evaluate our model in noisy con-texts, verifying that articulatory knowledge can provide benefits in adverse acoustic conditions. Third, we use a corpus of side-by-side speech and articulator trajectories to show that the HAMM can reasonably predict the movement of the articulators.
1. INTRODUCTION Hidden Markov Models are a popular method for speech recog-nition. Commonly, a left-to-right topology is used, where each phoneme is represented by a sequence of states, typically three [16]. In this topology, each state simply represents a portion of a phoneme. This acoustic-based model for speech recognition does not directly incorporate any knowledge of the source that pro-duced the speech.
In contrast, we know that speech is formed by the glottal excite-ment of a human vocal tract consisting of articulators which shape and modify the sound in complex ways. Since this system is limited by physical constraints, it could allow us to construct a more realistic model of speech to improve speech recognition. Such a model could have many advantages such as being better able to predict co-articulation effects, since they are due to physical limitations and energy-saving shortcuts in articulator movement [9]. Furthermore, by modeling articulators, we can allow asynchrony between their movements, which may more accurately model the production of speech [4]. Finally, because articulatory configurations are shared across multiple phonetic conditions, such a model may need less training data than a model without such information.
There has been much interest in incorporating articulatory knowledge into speech recognition. Gupta and Schroeter [8] discuss the analysis-by-synthesis approach, which attempts to
estimate the parameters of the Coker [3] model, which is based on articulatory features. The analysis-by-synthesis work is often targeted toward speech compression, where the quality of the synthesis is more important than the accuracy of the estimated parameters. The inverse mapping problem, the mapping of acoustic features to articulatory configurations, is discussed in [1]. In [10] Kirchhoff demonstrates how to use artificial neural networks to estimate articulatory features from acoustic features. The HAMM can also be cast as a factorial HMM [14] which has been attempted for speech recognition [11] without the use of articulatory knowledge. We chose to implement the HAMM using a standard HMM with a constrained state space equal to the Cartesian product of the components, as this allows us to use standard HMM algorithms for training and testing.
The rest of the paper is as follows: Section 2 presents the model and describes how it is initialized and trained. Section 3 presents experimental results.
2. THE MODEL A Hidden-Articulator Markov Model is based on the human articulatory system. Suppose we have N articulators in our model, and each articulator, a, can be in one of Ma positions. An articulatory configuration is an N-element vector C={c1,c2,…,cN}, where ca is an integer 0≤ca. A HAMM is an
HMM in which each hidden state represents a particular articula-tory configuration. The details of the model can be found in our previous work [13].
2.1 Articulatory Space A word is a sequence of articulator targets. In mapping words to articulator configurations, we make the simplifying assumption that words can be modeled as a sequence of phonemes, each of which is mapped to a sequence of one or more articulatory con-figurations.