An inverse design method for supercritical airfoil based on conditional generative models

Inverse design has long been an efficient and powerful design tool in the aircraft industry.In this paper, a novel inverse design method for supercritical airfoils is proposed based on generative models in deep learning. A Conditional Variational Auto Encoder(CVAE) and an integrated generative network CVAE-GAN that combines the CVAE with the Wasserstein Generative Adversarial Networks(WGAN), are conducted as generative models. They are used to generate target wall Mach distributions for the inve...     

空间发电系统热能储存容器的二维模型

高温熔盐相变储热是空间太阳能热动力发电系统的关键技术之一。基于微重力状态下的传热控制微分方程, 采用焓法对相变材料容器进行了二维数值分析。建立了一种新的空穴模型, 改进了以往的模拟计算, 对计算结果给予了讨论。      

(鱼参)科类鱼尾模型的巡游推进特性实验研究

通过对不同形状运动尾鳍模型的测力实验,研究了鱼 参科鱼类巡游时尾鳍的推进特性.结果表明:模型转动与平动间的相位差对模型受到推力的大小及方向有显著影响,在转动滞后平动90°附近模型受到较大推力.对于五种不同的尾鳍模型,虽然形状各异,但它们受到的平均推力正比于无量纲二阶矩与模型面积的乘积.     

一种基于滑动库的多模自适应控制方法

针对系统参数跳变时常规自适应控制闭环系统暂态响应差的问题,论文提出了一种基于滑动库的多模自适应控制方法。该控制器由多个参数已知的固定模型,一个自适应模型和一个滑动模型库构成。固定模型由被控对象在特定环境下的特征模型直接映射获得,其参数和其邻域一起覆盖被控对象的参数集;自适应模型对被控对象进行在线参数辨识,其参数被用来动态更新固定模型库;滑动模型库根据当前的工作点从固定模型库中选择离工作点较近的部分模型参与控制,优化控制模型集,减少计算量。仿真实验的结果表明了该方法的有效性。     

两种湍流模型及其可压缩修正在高超压缩拐角中的评估(英文)

采用在高阶精度差分格式对SA湍流模型和SST湍流模型及相应的可压缩修正模型在高超拐角流中进行了评估。可压缩修正方式考虑了密度梯度、压力膨胀和湍流马赫数等方法。为了减小数值误差与模型误差之间的混淆,控制方程的对流项采用了5阶精度的加权紧致非线性格式(WCNS-E-5),粘性项采用了一种半结点/结点交错的4阶中心格式。通过对马赫数为9.22的15度拐角和34度拐角湍流的模拟,考察了原始湍流模型及其修正模型的效果。计算表明:原始SST模型对高超拐角湍流的预测比原始SA模型准确,这种准确主要体现在对分离区预测、再附点附近压力和热流峰值预测上。通过混合采用Catris和Shur等的方法对 SA模型进行可压缩修正可以大大改进模拟效果。在SST模型的可压缩修正方法中,Catris的修正方法最好;考虑压力膨胀修正后得到的分离区远远偏大,比本原始SST模型更差;考虑湍流马赫数的修正方法得到的分离区偏小。本文还给了部分低阶格式的计算结果,高阶格式与低阶格式相比,对分离区大小、再附点附近的压力和热流峰值等的预测准度有所改进。     

测量液体推进剂剩余量的体积激励法

首先给出了以贮箱内气体为对象的一般热力学测量控制方程,完成气体及液体体积的计算; 以贮箱系统为研究对象,建立了系统处于等温、绝热及非等温条件下的热力学模型,考虑了气体与液体、贮箱壁的传热传质等因素,得出了比较理想的测量模型。其次,对影响测量的系统误差进行了分析,建立了各种影响因素的误差模型,并提出了相应的修正措施。然后,给出了在现有技术水平下测量的最大误差,结果表明测量误差在贮箱总体积的1%以内。最后,对激励频率、体积改变量大小及测量设备等相关应用技术问题进行了分析讨论与合理设计。体积激励法是一种可满足测量精度和耗能要求的良好方法。     

基于径向基神经网络的气动热预测代理模型

为快速获取高超声速飞行器表面热流数据并缩短飞行器气动热设计周期,提出了一种基于径向基神经网络的气动热快速预测代理模型方法。首先,在飞行器表面每一个离散化的网格节点单独构造一种正则化的径向基神经网络。随后,通过训练集对所有网络同时进行训练,获得各自网络的连接权值。最后,所有网格节点的神经网络协同预测飞行器表面不同位置的热流。对NASA火星实验室的椭圆钝化高超声速飞行器的应用表明,所提出的代理模型方法在模型训练完成后能够快速进行飞行器表面热流预测,并且模型具有良好的泛化能力,在驻点及迎风大面积区域热流预测结果与数值模拟的偏差在10%以内。     

Study of daytime vertical E × B drift velocities inferred from ground-based magnetometer observations of ΔH,at low latitudes under geomagnetically disturbed conditions

In this study, 30 storm sudden commencement (SSC) events during the period 2001–2007 for which daytime vertical E × B drift velocities from JULIA radar, Jicamarca (geographic latitude 11.91°S, geographic longitude 283.11°E, 0.81°N dip latitude), Peru and ΔH component of geomagnetic field measured as the difference between the magnitudes of the horizontal (H) components between two magnetometers deployed at two different locations Jicamarca (geographic latitude 11.91°S, geographic longitude 283.11°E, 0.81°N dip latitude) and Piura (geographic latitude 5.21°S, geographic longitude 279.41°E, 6.81°N dip latitude), in Peru, were considered. It is observed that a positive correlation exists between peak value of daytime vertical E × B drift velocity and peak value of ΔH for the three consecutive days of SSC. A qualitative analysis made after selecting the peak values of daytime vertical E × B drift velocity and ΔH showed that 57% of the events have daytime vertical E × B drift velocity peak in the magnitude range 20–30 m/s and 63% of the events have ΔH peak in the range 80–100 nT. The maximum probable (45%) range of time of occurrence of peak value for both vertical E × B drift velocity and ΔH during the daytime hours were found to be the same, i.e., 10:00–12:00 LT. A strong positive correlation was also found to exist between the daytime vertical E × B drift velocity and ΔH for all the three consecutive days of SSC, for all the events considered. To establish a quantitative relationship between day time vertical E × B drift velocity and ΔH, linear and polynomial (order 2 and 3) regression analysis (Least Square Method (LSM)) were carried out, considering the fully disturbed day after the commencement of the storm as ‘disturbed period’ for the SSC events selected for analysis. The formulae indicating the relationship between daytime vertical E × B drift velocity and ΔH, for the ‘disturbed periods’, obtained through the regression analysis were verified using the JULIA radar observed E × B drift velocity for 3 selected events. Root Mean Square (RMS) error analysis carried out for each case suggest that polynomial regression (order 3) analysis provides a better agreement with the observations from among the linear, polynomial (order 2 and 3) analysis.     

基于多模型切换的鲁棒自适应控制器

系统参数发生跳变时,为了解决多变量系统暂态响应变差的问题,提出一种基于多模型切换的鲁棒自适应控制器。该控制器由多个鲁棒自适应控制器组成,鲁棒自适应控制器保证系统存在未建模动态时能够使自适应系统稳定运行,引入的多模型策略保证系统参数跳变时具有良好的暂态性能。仿真结果表明,所提出的多模型鲁棒自适应控制器能够加快系统的响应速度,提高系统的暂态性能。     

DESIGN OBJECT MODEL AND DESIGN PROCESS MODEL OF COMPUTER AIDED AIRCRAFT CONCEPTUAL DESIGN SYSTEM

ＤＥＳＩＧＮＯＢＪＥＣＴＭＯＤＥＬＡＮＤＤＥＳＩＧＮＰＲＯＣＥＳＳＭＯＤＥＬＯＦＣＯＭＰＵＴＥＲＡＩＤＥＤＡＩＲＣＲＡＦＴＣＯＮＣＥＰＴＵＡＬＤＥＳＩＧＮＳＹＳＴＥＭＳｈｅｎＺｈａｎｇ（ＡｉｒｃｒａｆｔＣＡＤＲｅｓｅａｒｃｈＣｅｎｔｅｒ，Ｂｅｉｊｉｎ...