精子膜蛋白受精素(Fertilin)研究进展

精子膜蛋白受精素(Fertilin)是参与精卵质膜间相互作用最有特征性的候选分子之一。受精素由α、β两个亚基组成,为精子表面的异二聚体跨膜糖蛋白。fertilin-α(Fα)、Fertilin-β(Fβ)可能在精子发生、精子成熟、精子获能及精卵结合和融合中发挥重要作用。     

Dietary Fucoxanthin Increases Metabolic Rate and Upregulated mRNA Expressions of the PGC-1alpha Network,Mitochondrial Biogenesis and Fusion Genes in White Adipose Tissues of Mice

The mechanism for how fucoxanthin (FX) suppressed adipose accumulation is unclear. We aim to investigate the effects of FX on metabolic rate and expressions of genes related to thermogenesis, mitochondria biogenesis and homeostasis. Using a 2 × 2 factorial design, four groups of mice were respectively fed a high sucrose (50% sucrose) or a high-fat diet (23% butter + 7% soybean oil) supplemented with or without 0.2% FX. FX significantly increased oxygen consumption and carbon dioxide production and reduced white adipose tissue (WAT) mass. The mRNA expressions of peroxisome proliferator-activated receptor (PPAR) γ coactivator-1α (PGC-1α), cell death-inducing DFFA-like effecter a (CIDEA), PPARα, PPARγ, estrogen-related receptor α (ERRα), β3-adrenergic receptor (β3-AR) and deiodinase 2 (Dio2) were significantly upregulated in inguinal WAT (iWAT) and epididymal WAT (eWAT) by FX. Mitochondrial biogenic genes, nuclear respiratory factor 1 (NRF1) and NRF2, were increased in eWAT by FX. Noticeably, FX upregulated genes of mitochondrial fusion, mitofusin 1 (Mfn1), Mfn2 and optic atrophy 1 (OPA1), but not mitochondrial fission, Fission 1, in both iWAT and eWAT. In conclusion, dietary FX enhanced the metabolic rate and lowered adipose mass irrespective of the diet. These were associated with upregulated genes of the PGC-1α network and mitochondrial fusion in eWAT and iWAT.     

基于特征迁移学习的变工况下轴向柱塞泵故障诊断

不同工况下的轴向柱塞泵故障数据存在分布差异,现有的基于特征迁移学习的变工况故障诊断方法大多只通过单个传感器信号进行分析,具有一定的局限性和片面性。为了利用多传感器信号提高变工况下轴向柱塞泵故障诊断的性能,该研究提出一种耦合分类器子空间嵌入分布自适应（Subspace Embedded Distribution Adaptation with Coupled Classifiers, SEDACC）方法。该方法利用多传感器信号的频谱数据构造主要数据集和辅助数据集,通过子空间对齐（Subspace Alignment, SA）方法将源域和目标域的主要数据投影到公共子空间中,并采用加权条件最大均值差异（Weighted Conditional Maximum Mean Discrepancy, WCMMD）作为度量进行特征分布的适配。同时,基于结构风险最小化（Structural Risk Minimization, SRM）准则在源域标签数据上学习主分类器,根据主分类器对于目标域的预测结果在目标域辅助数据上学习辅助分类器。通过交替和迭代策略不断优化分类器参数,最后对二者进行加权融合得到最终的诊断模型。通过轴向柱塞泵变工况故障诊断试验进行验证,结果表明,当以垂直于端盖的z方向振动信号为主要数据并使用声音信号（或以平行于端盖的x方向的振动信号）作为辅助数据时,SEDACC方法在6种迁移任务中的平均准确率为99.88%（99.46%）,高于其他方法。此外,所提方法在目标工况样本稀少的情况下仍具有较高的诊断精度,当目标域和源域样本数比值为0.2时,6种迁移任务的平均准确率达到92.66%。研究结果可为更完备与准确的机械故障诊断提供参考。     

基于Transformer的强泛化苹果叶片病害识别模型

模型泛化能力是病害识别模型多场景应用的关键,该研究针对不同环境下的苹果叶片病害数据,提出一种可以提取多类型特征的强泛化苹果叶片病害识别模型CaTNet。该模型采用双分支结构,首先设计了一种卷积神经网络分支,负责提取苹果叶片图像的局部特征,其次构建了具有挤压和扩充功能的视觉Transformer分支,该分支能够提取苹果叶片图像的全局特征,最后将两种特征进行融合,使Transformer分支可以学习局部特征,使卷积神经网络分支学习全局特征。与多种卷积神经网络模型和Transformer模型相比,该模型具有更好的泛化能力,仅需学习实验室环境叶片数据,即可在自然环境数据下达到80%的识别精度,相较卷积神经网络EfficientNetV2提升7.21个百分点,相较Transformer网络PVT提升26.63个百分点,能够有效提升对不同环境数据的识别精度,解决了深度学习模型训练成本高,泛化能力弱的问题。     

多尺度分解双寻优策略SPCNN的果园苹果异源图像融合模型

针对单一的自然场景图像信息不能满足准确识别果实和精准定位目标的要求,提出一种多尺度分解双寻优策略简化脉冲耦合神经网络（Simplified Pulse Coupled Neural Network, SPCNN）的飞行时间（Time of Flight,ToF）与可见光果园苹果图像融合模型。对SPCNN模型引入带参数优化的双寻优策略,对非下采样轮廓波变换（Nonsubsampled Contourlet Transform, NSCT）融合规则进行改进。模型包括配准模块、编码区、多尺度分解模块、单目标SPCNN融合模型、多目标SPCNN融合模型、解码区。模型改进了SPCNN模型的参数优化方式以及迭代次数,模型自适应点火次数较低,在3～7次左右,具有点火次数低、自适应分割、效率高的优点。中光15:00时段点火识别成功率达到了100.00%,点火分割时间达到最低91.91s。与其他融合模型比较,模型在强光12:00、中光15:00、弱光18:20、19:00时段融合图像识别成功率达到100.00%;融合时间低于SPCNN模型,达到最低92.68s。模型识别精度最优达到了100.00%、融合耗时最低达到了92.68s、模型大小较SPCNN低一个数量级,可补充和完善图像层次融合理论和方法。     

Analysis of Spectral Characteristics Based on Optical Remote Sensing and SAR Image Fusion

Because of cloudy and rainy weather in south China, optical remote sens-ing images often can＇t be obtained easily. With the regional trial results in Baoying, Jiangsu province, this paper explored the fusion model and effect of ENVISAT/SAR and HJ-1A satel ite multispectral remote sensing images. Based on the ARSIS strat-egy, using the wavelet transform and the Interaction between the Band Structure Model （IBSM）, the research progressed the ENVISAT satel ite SAR and the HJ-1A satel ite CCD images wavelet decomposition, and low/high frequency coefficient re-construction, and obtained the fusion images through the inverse wavelet transform. In the light of low and high-frequency images have different characteristics in differ-ent areas, different fusion rules which can enhance the integration process of self-adaptive were taken, with comparisons with the PCA transformation, IHS transfor-mation and other traditional methods by subjective and the corresponding quantita-tive evaluation. Furthermore, the research extracted the bands and NDVI values around the fusion with GPS samples, analyzed and explained the fusion effect. The results showed that the spectral distortion of wavelet fusion, IHS transform, PCA transform images was 0.101 6, 0.326 1 and 1.277 2, respectively and entropy was 14.701 5, 11.899 3 and 13.229 3, respectively, the wavelet fusion is the highest. The method of wavelet maintained good spectral capability, and visual effects while improved the spatial resolution, the information interpretation effect was much better than other two methods.     

焊接温度对木塑窗扇热熔焊接强度的影响分析

利用四位热熔焊接机对木塑窗扇进行热熔焊接实验,所用窗扇型材是由36%的塑粉（PE）、60%的木粉及偶联剂、抗氧化剂等挤出而成。在焊接过程中,影响热熔焊接强度因素较多,本次试验重点考察焊接温度对窗角焊接的影响。在完成焊接后,通过对木塑窗扇角强度的测试,进行单因素方差分析,并观察焊接接口,结果证明温度对焊接强度具有显著的影响,在240~280℃范围内可实现该木塑窗型材的焊接,理想焊接温度为260℃左右。     

基于多元数据融合的土地生态指数构建——以金坛市为例

以金坛市为研究对象,根据该市区域特点选取指标、确定数据来源和计算方法、进行标准化函数及权重确定,得到了金坛市行政村尺度的土地生态指数。分布规律表明金坛市土地生态质量东部高于西部,以金坛主城区为中心向四周递增的圈层状发展趋势大致符合城乡梯度规律。     

峨山彝语地名的文化语言学解读

峨山自古就是彝族聚居的地方,传承了众多的彝语地名,其命名与彝族所处自然环境、社会生活和历史文化密切相关,体现出明显的区域特征和民族特征,从中可以看到峨山彝族社会历史风貌、语言文化和多民族文化的交流与融合.     

融合多源图像信息的果实识别方法

光线变化与目标重叠是影响自然环境中果实正确识别的重要原因。为了降低两者的影响,研究了融合多源图像信息的果实识别方法。在图像配准的基础上,优选了H分量图与幅度图像作为待融合的源图像;由模糊推理系统(隶属度函数和模糊规则)决定权重,采用加权平均策略实现图像的像素级融合;根据融合图像中果实区域的分布规律,设计了一种基于直方图的首阈检测法以获得最佳的果实分割效果;利用深度图像的统计特性,设计了一种逐层分割图像的方法以解决重叠果实的分离问题。实验结果表明:多源融合图像用于果实识别与定位比单一图像具有更好的准确性与鲁棒性,对重叠果实的正确识别率在83.67%~94.22%之间。