基于D-ELM的矿井气体FTIR光谱定量分析

为了对D-ELM的矿井气体定量分析效果进行比较和分析,建立了3种矿井气体定量分析模型,分别是SVM模型、极限学习机（ELM）模型和动态极限学习机（D-ELM）模型。由于每个模型每次预测结果在一定范围内变化,所以取每个模型运行10次预测结果平均均方根误差（ARMSE）、平均相关系数（AR）、平均模型运行时间（AT）来评价各模型对气体定量分析的性能。PSO-SVM模型的结果为:[0.054 2,0.998,200.38 ];ELM模型的结果为:[1.042 1,0.989 4,0.26];D ELM模型结果为:[0.043 8,1,2.01]。综合预测精度和预测速度表明,D ELM模型要优于另外2种模型。     

Macroscopic-microscopic calculations of ground state properties of superheavy nuclei

We systematically calculate the ground state properties of superheavy even-even nuclei with proton number Z=94–118. The calculations are based on the liquid drop macroscopic model and the microscopic model with the modified single-particle oscillator potential. The calculated binding energies and α-decay energies agree well with the experimental data. The reliability of the macroscopic-microscopic(MM)model for superheavy nuclei is confirmed by the good agreement between calculated results and experimental ones. Detailed comparisons between our calculations and M?ller’s are made. It is found that the calculated results also agree with M?ller’s results and that the MM model is insensitive to the microscopic single-particle potential. Calculated results are also compared with results from relativistic mean-field (RMF) model and from Skyrme-Hatree-Fock(SHF) model. In addition, half-lives, deformations and shape coexistence are also investigated. The properties of some unknown nuclei are predicted and they will be useful for future experimental researches of superheavy nuclei.     

爆磁压缩发生器的爆炸管动力学效应

 对爆磁压缩发生器中爆炸管2维动力学简化模型进行了模拟计算，分析表明：径向膨胀速度会随径向位置（或者时间）的变化而变化，因此膨胀角也会随径向位置（或者时间）发生变化。对各时间点（或位置点）处的膨胀速度进行了平均，求得理论上的平均膨胀速度，再将该平均膨胀速度与实验测量值进行了比较。模拟结果给出了径向膨胀速度受到端头效应影响的情况，这可为改进实验结果提供参考。由于径向速度与轴向速度的比值一般在5以上，用作爆炸管的物质质量越大，这一比值就越大，因此选择密度较大的金属材料作为爆炸管，可减少滑移。应用2D简化模型计算出的膨胀角数值，与Gurney模型以及1D模型进行了比较，它们之间的差别可能主要来自2D效应。     

基于中红外光谱技术的香菇蛋白质含量测定

研究了中红外光谱预测香菇蛋白质含量的可行性。去掉明显噪声部分后,研究香菇3 581～689 cm-1中红外光谱与蛋白质含量的关系。以Savitzky-Golay(SG)5点平滑预处理光谱建立偏最小二乘法(partial least squares, PLS)的预测模型的效果不理想,模型的建模集和预测集的相关系数均高于0.85,但剩余预测偏差(residual prediction deviation, RPD)值仅为1.77。采用连续投影算法(successive projections algorithm, SPA)算法从3000个波数点中选择7个特征波数,并以七个特征波数分别建立PLS、多元线性回归(multiple linear regression, MLR)、反向传播神经网络(back-propagation neural network, BPNN)和极限学习机模型(extreme learning machine, ELM)。与全谱的PLS相比,以特征波数的PLS模型和MLR模型的预测效果相对较差,而以特征波数的BPNN和ELM模型的预测效果相对较好。其中SPA-ELM模型的预测效果最佳,预测集相关系数(correlation coefficient of prediction)R_p=0.899 5,预测集均方根误差(root mean square error of prediction)RMSEP=1.431 3,剩余预测偏差RPD=2.18。研究结果表明,中红外光谱分析技术可以用于预测香菇蛋白质含量,且SPA选取特征波数能用来代替原始光谱进行建模分析,为香菇蛋白质含量的检测提供了新的思路。     

Novel model of a AlGaN/GaN high electron mobility transistor based on an artificial neural network

In this paper we present a novel approach to modeling AlGaN/GaN high electron mobility transistor(HEMT) with an artificial neural network(ANN).The AlGaN/GaN HEMT device structure and its fabrication process are described.The circuit-based Neuro-space mapping(neuro-SM) technique is studied in detail.The EEHEMT model is implemented according to the measurement results of the designed device,which serves as a coarse model.An ANN is proposed to model AlGaN/GaN HEMT based on the coarse model.Its optimization is performed.The simulation results from the model are compared with the measurement results.It is shown that the simulation results obtained from the ANN model of AlGaN/GaN HEMT are more accurate than those obtained from the EEHEMT model.     

Application of the Englert-Schwinger model to the equation of state and the fullerene molecule

The Englert-Schwinger model (ESM) is applied to two problems. One is the calculation of zero-temperature equation of state (EOS) of elements within the spherically symmetric Wigner-Sietz cell approximation. The other is to obtain the equilibrium radius of fullerene molecule using March’s approach [N. H. March, Proc. Camb. Philos. Soc. 48, 665 (1952)]. In each case, the results of the ESM are compared with those of Thomas-Fermi-Dirac (TFD) and Thomas-Fermi-Dirac-Weizsacker (TFDW) models. Zero-temperature equation of state calculations are done for Al and Cu. The results of the ESM show an enormous improvement over those of the TFD model. Also, the ESM is in good agreement with the TFDW model for compressions greater than 2. In the regime of validity of TFDW theory, i.e., compressions greater than 20 and 10 for Al and Cu, respectively, the deviations between the results of the two models are negligible. Hence, the ESM may be used in lieu of the TFDW model for EOS calculations. In the fullerene case, we have obtained the cohesive energy using the models assuming the radius obtained from accurate calculations of the fullerene molecule. We have also obtained the equilibrium radius predicted by each model. The results obtained show that the ESM results are not much of an improvement over the TFD results. This shows that the ESM cannot always improve the results of the TFD model and be a replacement for the TFDW model. However, as in the EOS case, it would give results in good agreement with TFDW results for properties that are dependent on the electron density at the outer reaches of the atom.     

Modeling of Focused Acoustic Field of a Concave Multi-annular Phased Array Using Spheroidal Beam Equation

A theoretical model of focused acoustic field for a multi-annular phased array on concave spherical surface is proposed.In this model,the source boundary conditions of the spheroidal beam equation(SBE) for multi-annular phased elements are studied.Acoustic field calculated by the dynamic focusing model of SBE is compared with numerical results of the O'Neil and Khokhlov-Zabolotskaya-Kuznetsov(KZK) model,respectively.Axial dynamic focusing and the harmonic effects are presented.The results demonstrate that the dynamic focusing model of SBE is good valid for a concave multi-annular phased array with a large aperture angle in the linear or nonlinear field.     

无氧铜动态卸载行为的数值模拟

 分别利用Johnson-Cook（JC）本构模型和Steinberg-Cochran-Guinan（SCG）本构模型对无氧铜在10~20 GPa冲击压力下的卸载过程进行了数值模拟,与冲击-卸载实验结果比较表明,在这样的冲击压力下无氧铜的应变率效应仍然明显,JC模型对无氧铜的动态卸载行为有较好的描述。     

酰肼类抗结核病药物的3D-QSAR研究

采用三维全息原子场作用矢量(3D-HoVAIF)对115个酰肼类抗结核病药物进行定量构效关系(QSAR)研究。运用多元线性回归(MLR)和偏最小二乘回归(PLS)建模,同时采用内部及外部双重验证的办法对所得模型稳定性能进行了深入分析和检验。MLR建模和PLS建模的复相关系数(Rcum2)、留一法交互校验复相关系数(QLOO2)以及外部样本校验复相关系数(Qext2)分别为0.765、0.663、0.748和0.733、0.614、0.715。结果表明,3D-HoVAIF能较好表征抗结核药物分子结构信息,所建QSAR模型具有良好稳定性和预测能力,为抗结核病药物的研发提供一定的理论基础。     

A new mixed subgrid-scale model for large eddy simulation of turbulent drag-reducing flows of viscoelastic fluids

A mixed subgrid-scale(SGS) model based on coherent structures and temporal approximate deconvolution(MCT) is proposed for turbulent drag-reducing flows of viscoelastic fluids. The main idea of the MCT SGS model is to perform spatial filtering for the momentum equation and temporal filtering for the conformation tensor transport equation of turbulent flow of viscoelastic fluid, respectively. The MCT model is suitable for large eddy simulation(LES) of turbulent dragreducing flows of viscoelastic fluids in engineering applications since the model parameters can be easily obtained. The LES of forced homogeneous isotropic turbulence(FHIT) with polymer additives and turbulent channel flow with surfactant additives based on MCT SGS model shows excellent agreements with direct numerical simulation(DNS) results. Compared with the LES results using the temporal approximate deconvolution model(TADM) for FHIT with polymer additives, this mixed SGS model MCT behaves better, regarding the enhancement of calculating parameters such as the Reynolds number.For scientific and engineering research, turbulent flows at high Reynolds numbers are expected, so the MCT model can be a more suitable model for the LES of turbulent drag-reducing flows of viscoelastic fluid with polymer or surfactant additives.     

Tube vibration in a half-scale sector model of a helical tube steam generator

This paper presents the experimental technique and results of tests on a half-scale sector model of a steam generator helical coil tube bank. A series of tests was performed: (1) bench tests of a single helical tube in air; (2) tests of the sector model in air; (3) tests of the sector model in stationary water to determine natural frequencies and damping; (4) tests in flow. The experimental results reveal the general characteristics of the sector model and provide the information for the design evaluation of a helical tube array to avoid detrimental fluidelastic instability.     

RANS predictions of turbulent diffusion flames: comparison of a reactor and a flamelet combustion model to the well stirred approach

The flame stabilisation process in turbulent non-premixed flames is not fully understood and several models have been developed to describe the turbulence–chemistry interaction. This work compares the performance of the multiple Representative Interactive Flamelet (mRIF) model, the Volume Reactor Fraction Model (VRFM), and the Well-Stirred reactor (WS) model in describing such flames. The predicted ignition delay and flame lift-off length of n-heptane sprays are compared to experimental results published within the Engine Combustion Network (ECN). All of the models predict the trend of ignition delay reasonably well. At a low gas pressure (42 bar) the ignition delay is overpredicted compared to the experimental data, but the difference between the models is not significant. However, the predicted lift-off lengths differ. At high pressure (87 bar) the difference between the models is small. All models slightly underpredict the lift-off length compared to the experimental data. At low gas pressure (42 bar) the mRIF model gives the best results. The VRFM and WS models predict excessively short lift-off lengths, but the VRFM model gives better results than the WS model. The flame structures of the models are also compared. The WS model and the VRFM model yield a well defined flame stabilisation point whereas the mRIF model does not. The flame of the mRIF model is more diffuse and the model is not able to predict flame propagation. All models were able to predict the experimental trends in lift-off and ignition delay, but certain differences between them are demonstrated.     

带支撑装置的CRM-WBH模型流固耦合数值模拟

共同研究模型（common research model，CRM）机翼/机身/平尾组合体构型（CRM-WBH）是第4届AIAA阻力预测研讨会（Drag Prediction Workshop IV，DPW IV）选择的基准构型，DPW IV会议的统计分析结果显示计算结果与试验结果之间存在显著差异.采用CFD方法和流固耦合方法数值模拟了带支撑装置的CRM机翼/机身/平尾组合体模型（CRM-WBHS）的气动特性，主要目的是评估支撑装置和静气动弹性变形对CRM-WBH模型气动特性数值模拟结果的影响.通过与CRM-WBH模型CFD数值模拟结果和欧洲ETW风洞（European Transonic Wind Tunnel）测力、测压和模型变形测量结果的对比分析，表明模型支撑装置导致机翼上翼面激波位置前移，升力系数、阻力系数下降，俯仰力矩系数增加；静气动弹性变形主要影响机翼上表面激波位置和外翼处激波位置前负压，导致升力系数、阻力系数进一步下降，俯仰力矩系数进一步增加.CRM-WBHS模型的流固耦合数值模拟结果更加接近试验结果.      

在40MeV/u时112Sn+112Sn和124Sn+124Sn中多重碎裂的同位旋效应

利用同位旋相关的量子分子动力学模型,研究了112Sn+112Sn和124Sn+124Sn两个反应系统在入射能量E=40MeV/u时的多重碎裂.计算结果能与实验值定性符合.观察到两个反应系统中,中等质量碎片多重性、中子多重性、荷电粒子多重性与轻荷电粒子多重性之间的关联存在着明显的差别.另外,通过与膨胀蒸发源模型及同位旋相关的渗透模型分析结果的比较,发现这种差别主要是由同位旋相关的反应动力学所造成的.     

三维描述子用于咪唑衍生物类抗艾滋病药物QSAR研究

采用三维全息原子场作用矢量(3D-HoVAIF)对42个5-pheny-l-phenylamino-1H-imidazole 类抗艾滋病药物进行定量构效关系(QSAR)研究。运用偏最小二乘回归(PLS)建模,同时采用内部及外部双重验证的办法对所得模型稳定性能进行了深入分析和检验。PLS建模的复相关系数、留一法交互校验复相关系数和外部样本校验复相关系数分别为0.912、0.847、0.931。结果表明,3D-HoVAIF能较好表征5-pheny-l-phenylamino-1H-imidazole 类抗艾滋病药物分子结构信息,因而能建立具有良好稳定性和预测能力的QSAR模型,为抗艾滋病药物的研发提供一定的理论基础。     

在二带京兹堡朗道理论下MgB2的超导参量的研究

文中通过G-L理论二带模型来研究相干长度ξ(T)、伦敦穿透深度λ(T)以及京兹堡朗道参量κ(T)=λ(T)/ξ(T)在Tc附近对温度T的依赖关系。理论计算结果与MgB2样品的实验数据符合的很好。同时与单带模型算出的结果做了对比,发现G-L理论二带模型能更精确描述MgB2的超导特性。     

初始模型对含噪动态光散射数据正则化反演结果的影响

分别采用最小模型矩阵、最平坦模型矩阵、最光滑模型矩阵作为初始化模型,对加入5种不同水平随机噪声的90nm窄单峰、90nm宽单峰和250nm窄单峰、250nm宽单峰颗粒体系的模拟分布进行了正则化反演,并对反演结果进行比较。结果表明:当噪声水平为0时,正则化初始模型的选择对反演结果没有明显影响。随着噪声水平的增加,采用三种初始化模型反演得到的峰值误差和粒度分布误差都随之变大,但采用最平坦模型和最光滑模型反演得到的峰值和粒度分布误差明显小于采用最小初始模型的反演误差。当噪声水平大于0.01时,选择最平坦初始模型获得的粒度分布结果优于采用最光滑初始模型和最小初始模型获得的结果,而采用最光滑初始模型反演得到的峰值优于最平坦初始模型和最小初始模型的反演峰值。因此,采用正则化算法处理含噪动态光散射数据时,为得到最优的粒度分布信息,宜采用最平坦初始模型,若需要获取最准确的峰值信息,则应选择最光滑初始模型。     

旋流燃烧NO生成的USM湍流反应模型

用统一二阶矩(USM)湍流反应模型对不同旋流数下甲烷-空气旋流燃烧NO生成进行了数值模拟,并和EBU-Arrhenius(E-A)燃烧模型对燃烧的模拟结果和简化PDF湍流反应模型对NO生成的模拟结果以及相应的实验结果进行对比。结果表明,USM模型显著地优于E-A模型和简化PDF模型。E-A模型不能合理地模拟有限反应动力学,而简化PDF模型用两个单变量PDF的乘积代替联合PDF,大大地低估了NO湍流反应率。USM模型预报结果和实验结果符合最好。     

Effect of viscosity on stability and accuracy of the two-component lattice Boltzmann method with a multiple-relaxation-time collision operator investigated by the acoustic attenuation model

A two-component lattice Boltzmann method (LBM) with a multiple-relaxation-time (MRT) collision operator is presented to improve the numerical stability of the single relaxation time (SRT) model. The macroscopic and the momentum conservation equations can be retrieved through the Chapman—Enskog (C-E) expansion analysis. The equilibrium moment with the diffusion term is calculated, a diffusion phenomenon is simulated by utilizing the developed model, and the numerical stability is verified. Furthermore, the binary mixture channel model is designed to simulate the sound attenuation phenomenon, and the obtained simulation results are found to be consistent with the analytical solutions. The sound attenuation model is used to study the numerical stability and calculation accuracy of the LBM model. The simulation results show the stability and accuracy of the MRT model and the SRT model under different viscosity conditions. Finally, we study the influence of the error between the macroscopic equation of the MRT model and the standard incompressible Navier—Stokes equation on the calculation accuracy of the model to demonstrate the general applicability of the conclusions drawn by the sound attenuation model in the present study.     

Size effect of the elastic modulus of rectangular nanobeams:Surface elasticity effect

The size-dependent elastic property of rectangular nanobeams （nanowires or nanoplates） induced by the surface elas- ticity effect is investigated by using a developed modified core-shell model. The effect of surface elasticity on the elastic modulus of nanobeams can be characterized by two surface related parameters, i.e., inhomogeneous degree constant and surface layer thickness. The analytical results show that the elastic modulus of the rectangular nanobeam exhibits a distinct size effect when its characteristic size reduces below 1 O0 nm. It is also found that the theoretical results calculated by a mod- ified core-shell model have more obvious advantages than those by other models （core-shell model and core-surface model） by comparing them with relevant experimental measurements and computational results, especially when the dimensions of nanostructures reduce to a few tens of nanometers.