基于小波阈值降噪和经验模态分解的高光谱图像分类算法

为了充分降低高光谱图像中的噪声以获得高精度的分类结果,本文结合小波阈值降噪(WTD)和经验模态分解(EMD)的优点,提出了一种基于小波阈值降噪-经验模态分解的高精度支持向量机(SVM)高光谱图像分类算法(WTD-EMD-SVM)。首先对高光谱图像进行小波阈值降噪,除去高光谱数据中的高频噪声;然后再对高光谱图像进行EMD,获得含有高光谱数据本质特征的内固模态函数(IMF)和含有低频噪声的残差;最后采用内固模态函数重构高光谱图像,并对高光谱图像进行SVM分类。将其应用到AVIRIS数据92AV3C,仿真结果表明该算法不仅提高了高光谱图像分类精度,同时可减少支持向量数目,以提高高光谱图像分类速度。     

最优特征选择下多层次分割的城市道路提取

针对道路提取过程中特征维数过高的问题,提出了一种基于ReliefF过滤式和Wrapper封装式的特征选择方法.将粒子群优化算法(PSO)作为Wrapper的搜索算法,优化过的随机森林算法(OPRF)作为Wrapper的分类器构成PSO_OPRF封装式子集评估器,对ReliefF预选后的特征子集进行评估,降低特征维度,选...     

Transfer learning: A new aerodynamic force identification network based on adaptive EMD and soft thresholding in hypersonic wind tunnel

The aerodynamic test in the pulse combustion wind tunnel is very important for the design, evaluation and optimization of aerodynamic characteristics of the hypersonic aircraft. The test accuracy even affects the success or failure of hypersonic aircraft development. In the aerodynamic test of pulse combustion wind tunnel, the aerodynamic signal is disturbed by the inertial force signal, which seriously affects the test accuracy of aerodynamic force. Aiming at the above problems, this paper innovatively proposes an aerodynamic intelligent identification method, that is the transfer learning network based on adaptive Empirical Modal Decomposition (EMD) and Soft Thresholding (TLN-AE&ST). Compared with the existing aerodynamic intelligent identification model based on deep learning technology, this study introduces the transfer learning idea into the aerodynamic intelligent identification model for the first time. The TLN-AE&ST effectively alleviates the problem of scarcity of training samples for intelligent models due to the high cost of wind tunnel tests, and provides a new idea for further implementation of deep learning technology in the field of wind tunnel aerodynamic testing. And this study designed residual attention block with soft threshold and dense block with adaptive EMD in TLN-AE&ST model. Residual attention block with soft threshold module can more effectively suppress the influence of instrument noise signal on model training effect. Dense block with adaptive EMD makes the deep learning model no longer a black box to a certain extent, and has certain physical significance. Finally, a series of wind tunnel tests were carried out in the Φ = 2.4 m pulse combustion wind tunnel of China Aerodynamic Research and Development Center to verify the effectiveness of TLN-AE&ST.     

脉冲耦合神经网络在图像分割中的应用研究

系统研究了脉冲耦合神经网络在图像分割中的应用。首先研究了标准PCNN和简化的PCNN模型在图像分割中的应用,对分割结果进行了比较分析,结果表明,两种PCNN模型都能得到较好的分割结果。其次研究了最大熵、最小交叉熵和区域特征一致性准则在PCNN最优分割结果确定中的应用。最后探讨了PCNN参数变化对图像分割结果的影响,并给出了仿真试验结果。     

改进的SSDA并行算法及其在TEM中的应用

介绍了一种把图像匹配中的序贯相似性检测算法(SSDA,Sequential Similarity Detection Algorithm)应用到地形匹配中的方法,把地形匹配系统中的数字高程模型、高程数据和实时剖面数据分别视为SSDA算法中的搜索图、灰度值和模板,同时提出了动态门限序列和分组选取随机点的方法,最后将改进的SSDA算法进行并行化并用消息传递接口(MPI, Message Passing Interface)实现.实验结果表明,改进的SSDA算法能有效地提高地形匹配的速度和计算精度,而相应的并行程序也能取得较好的加速比,从而一定程度上解决了传统的地形匹配算法时间复杂度高、实时性差的问题.     

基于分水岭与图割的自动分割方法

为快速、准确的提取CT序列图像中目标物体,把分水岭和图割相结合.首先选择目标物体的内外轮廓,对内外轮廓之间的区域用分水岭算法预分割为若干小区域,把每一个小区域作为一个节点,建立图结构.把多源点和多汇点简化成单源点和单汇点,建立新的图结构.然后利用最大流/最小割定理进行切割,提取目标物体.最后把上一张CT目标物体的轮廓映射到下一张CT上,分别扩大和缩小该轮廓作为该CT的内外轮廓.根据上述方法提取轮廓,对整个CT序列依次循环操作.通过实验证明该算法在分割效果和分割时间上优于其它传统算法,同时,实现了三维空间上序列轮廓的自动提取.     

航空复杂产品的 CATIA 结构稳健建模方法研究

针对现有建模方法无法有效对航空产品 CATIA 模型进行快速修改与重用的问题,提出了以曲面分割为建模策略的结构稳健性建模方法。该方法通过对父子关系关联传播路径进行分析,找出影响 CATIA 特征模型结果稳健性的因素,即几何元素方向、几何拓扑元素、建模策略。通过研究特征构建时,输人元素的性质和基于此性质的操作方法,实现结构稳健性建模,即通过对外部输人和建模过程中的关键几何元素进行方向控制、创建显式参考代替几何拓扑元素、禁止使用复杂草图、采用基于曲面分割的设计特征模块化建模策略等方式,实现特征模型的快速修改与重用。工程实例证明了所提方法的实用性和可靠性。     

航空发动机双转子系统临界转速求解方法

针对航空发动机转子动力学设计需求,将有限元法与振型筛选法相结合,提出适于工程应用的临界转速求解方法,结合算例详细论述了完全法和缩减法的求解过程。研究结果表明:两种求解方法的结果完全一致,而缩减法求解过程迭代次数少,更为方便;某双转子系统,顺转时存在5阶临界转速,而反转时在相同的转速范围内将增加n=3,n=4和n=5的反进动临界转速,临界转速的密集性导致反转转子的动力学设计更为复杂。     

双S弯喷管的流固耦合特性研究

为了明晰S弯喷管的流固耦合特性,数值模拟了流固耦合作用下双S弯收敛喷管的结构变形特征及其内/外流特性。结果表明：S弯喷管的圆转方弯曲构型产生了非均匀的流场分布,并增强了结构的弹性特征,它们通过交换气动载荷与变形位移数据形成了S弯喷管流固耦合的作用机理。在气动载荷作用下,S弯喷管沿Y向的最大变形位移为25.3mm,它位于喷管出口上壁面的中心位置。当喷管的结构变形稳定时,第二弯转弯处下壁面的气流加速至局部超音速,壁面静压大幅降低;第一弯下游上壁面附近形成了气流分离区;喷管出口喷流沿轴向向上偏转。流固耦合作用导致S弯喷管的流量系数减小0.6%,推力系数降低1.8%。矩形截面与弯曲构型是影响S弯喷管流固耦合特性的主要几何特征,其中矩形截面能够显著增大喷管的变形位移;S形弯曲构型虽然能够有效抑制变形特征,但它导致喷管多个区域出现变形,喷管的结构变形分布变得更加复杂。     

Optimization for aerodynamic performance of double serpentine nozzles with spanwise offsets using Taguchi-based CFD analysis

Serpentine nozzles are widely used in combat aircraft to realize strong stealth characteristics. Based on the layout characteristics within a confined space, a series of double serpentine nozzles with spanwise offsets are established. Using computational fluid dynamics and Taguchi method, the influence mechanisms of the Distribution of Area (DA), Distributions of Centerline for the first and second ‘S’ sections in the Vertical direction (DCV1 and DCV2), and Distribution of Centerline in the Spanwise direction (DCS) are analyzed. The impact of these factors on the total pressure recovery coefficient can be ranked as DA > DCV2 > DCS > DCV1, whereas their impacts on the discharge coefficient and axial thrust coefficient can be ranked as DCV2 > DCS > DA > DCV1. Considering the statistical significance of these factors, a nozzle in which DA changes rapidly at the exit and DCV1, DCV2, and DCS change rapidly at the entrance gives the best aerodynamic performance. Compared to the worst configuration, the total pressure recovery coefficient, discharge coefficient, and axial thrust coefficient are improved by 1.6%,3.5% and 3.6%, respectively. DA influences the gas flow acceleration in the entire serpentine channel, resulting in different wall shear stress and friction losses. The various centerline distributions influence the gas flow acceleration effects and form complex wave structures in the constant-area extension section, resulting in different local and friction losses.