基于控制点影像库的合成孔径雷达图像几何校正

近年来,多谱段和高分辨率合成孔径雷达(SAR)成像技术已取得巨大进步,人们也高度重视其潜在的应用前景.但是,因为雷达斜视成像性质和地形起伏影响,SAR图像在山区的几何畸变非常复杂,采用常规方法不能达到正射校正的结果,所以目前SAR图像的应用情况并不尽人意.实际应用的雷达图像正射纠正方法,特别是密集三角形格网的数字微分纠正方法需要大量纠正控制点,但是,雷达图像噪声和目标边缘模糊使自动匹配的选点算法难以成功.提出了一种基于控制点影像库的纠正控制点选点方法,用以提高点位精度和工作效率.该方法包括纠正区域控制点影像建库、灵活交互操作机制设计和点匹配操作.对不同空间分辨率雷达图像的实验处理结果表明,该方法具有较好的应用价值.     

汕尾市100kW波力电站空气透平与气室匹配设计

根据用1/15模型,在造波水槽中进行试验获得气室的Δ Pi- Q、ηA- Q、 NA- Q特性,以及用300mm透平模型,在1000mm活塞式往复流透平试验台进行试验获得的对称翼透平、双向固定导叶冲动透平和双向自调节导叶冲动透平的CP-、ηT-特性,进行了气室与三种空气透平的匹配设计。表明在设计波况下,采用1.025m的双向自调节导叶冲动透平时,ηT·ηA达50.43%;采用0.975m的双向自调节导叶冲动透平时,ηT·ηA达46.66%;采用0.975m的双向固定导叶冲动透平时,ηT·ηA达33.66%;采用1.45m对称翼透平时,ηT·ηA仅为29.9%。并对四种不同匹配方案进行了对比分析。     

关于Barban定理的推广

本文将Barban定理中和式形式加以推广,得到关于函数μ(n)的这种和式的估计。     

I(L)值完全诱导空间

作者引入 I(L)值完全下半连续映射 ,研究其性质。利用 I(L)值完全下半连续映射定义I(L)值完全诱导空间 ,给出 I(L)值完全诱导空间的拓扑基的表达形式 ,证得两个 I(L)值完全诱导空间的映射是连续映射的充分必要条件 ,并建立了乘积空间的 I(L)值完全诱导空间与 I(L)值完全诱导空间的乘积空间的联系     

Diagnosis of iridovirus in large yellow croaker by PCR

A rapid and sensitive PCR-based method for the detection of the large yellow croaker iridovirus(LYCIV) is described,which involves the amplification of a 295 bp fragment of the LYCIV ATPase gene from DNA isolated from naturally infected fish spleen.Sequencing of LYCIV ATPase gene fragment showed it shared 100% nucleotide sequence homology with the corresponding region of the ATPase gene of red sea bream iridovirus(RSIV) and sea bass iridovirus(SBIV),suggesting that LYCIV was homologous with RSIV and SBIV at least in part of the gemone.The specificity and sensitivity of the PCR procedure were tested on the iridovirus-infected fishes,the expected fragment was detected from spleen DNA samples of infected fishes,whereas no fragments were amplified from healthy fish spleen DNA,white spot syndrome baculoviruses(WSBV) DNA and pseudorabies virus(PRV) DNA.Detection limit of this method was 10^-7 ng positive plasmid DNA containing target sequence,equal to about 100 virions.In the infected experiment,first positive detection(1/4) appeared at Day 3 post-infection,all fish(4/4) tested positive at Day 7,however obvious symptoms were observed at Day 8,so LYCIV infection could be detected prior to the appearance of obvious symptoms.These results indicate that this PCR method could be used for early,rapid and specific detection of LYCIV infection.     

基于EMD与神经网络的机械故障诊断技术

经验模式分解 (EMD)是分析非线性、非平稳信号的有力工具 ,它将信号分解为突出了原信号的不同时间尺度的局部特征信息的内在模函数 (IMF)分量。本文通过将各 IMF分量输入到 BP网络中进行训练学习和故障诊断 ,比直接输入原信号可以提高 BP网络对故障诊断的准确率 ,而且减少了训练时间。     

ECOM模式在丁字湾的应用

应用河口、陆架和海洋模式 (ECOM模型 ) ,引入干湿网格法模拟潮滩涨落的改进 ,并建立丁字湾及近岸海域的三维变动边界潮流模型。该模型考虑了湾口拦门沙、湾内水道和人工围海等地形特点。计算结果与实测值比较符合良好 ,较好地刻画出丁字湾 M2 分潮潮流场的时空分布特点。     

长期非平稳荷载调谐质量减振阻尼器优化设计

研究外荷载为长期非平稳随机过程。考虑长期荷载的特性 ,采用 1个概率谱密度函数来反映长期非平稳随机荷载及其特征 ;概率谱密度函数是基于大量的一般谱密度函数的统计特性获得。以延长结构的抗疲劳使用寿命为目标函数 ,提出了调谐质量减振阻尼器的优化设计方法 ,这在实际工程中有着极为广阔的应用前景。本文旨在从理论上发展长期非平稳随机荷载作用下调谐质量减振阻尼器的优化设计方法 ;文中采用长期波浪实测数据 ,给出了 1个数值算例说明整个设计过程。     

潮汐资料的统计分析（二）

4．1 随机差分方程模式的建立前节所建立的潮汐资料统计模式(4)要求相继的潮位必须是相互独立，如果实际观测表明并非如此，譬如上一时刻的潮位高，下一时刻的潮位平均地说也高，则模式(4)自然不太合适。为此，我们将引入一个比模式(4)更一般的潮汐观测资料统计模式，即下面所谓随机差分方程模式：     

Responses of CO2 in surface water in subtropical West Pacific to El Niño event

In order to investigate the relationships between the change of TCO2, △PCO2 and SST, current, upwelling and biological activities during El Nin(～)o event in the subtropical Pacific, the responses of TCO2 and △PCO2 in surface water in the subtropical Pacific during El Nin(～)o and La Nina have been simulated using a three-dimension carbon cycle model with biota pump. The results of numerical simulations show that TCO2 in sea water increases with reducing of SST during mature phase of El Nin(～)o in the subtropical West Pacific . At the same period , the Kuroshio in this region was weakened , the zonal currents were divergence , the upwelling carried the water with high concentrations of CO2 to the sea surface , so both of TCO2 and △PCO2 in surface water were increased . But TCO2 and △pCO2 were decreased during La Nina period. These simulated results confirmed the observations in 1982/1983 , 1986/1987 ,1991/1995 and 1997/1998 El Nin(～)o events .