基于PIV技术不同形式防波堤周围涡旋场特性研究(英文)

深水防波堤周围流场存在复杂的紊动变化,防波堤较容易发生失效破坏。本文利用PIV技术,基于模型实验,对规则波作用下防波堤周围的速度流场及其涡流特性进行了实验研究,获得了规则波流场的瞬时速度分布。通过对不同型式防波堤在不同的水力条件下的波浪流场实验数据的分析,定量地研究了防波堤堤身水域的涡旋流场的水力特性。     

基于特征区域统计的多光谱遥感影像海岛瞬时岸线提取(英文)

本研究采用IKONOS遥感图像,选取波谱特征区,通过区域像元统计并计算各波段权重系数,将影像进行波谱归一化处理,使归一化的影像岛陆与海水特征更明显。在此基础上,采用最大类间方差法确定最佳分割阈值,对归一化的影像进行二值化,从而提取海岛岸线。该方法采用特征曲线法进行水陆分离,得到的二值图像保持了原图的有效边缘,采用二值形态学提取的海岸线连续可靠、信噪比高。结果表明,该方法简单、快速,能有效提取海岸线,具有实用价值。     

基于GA参数辨识的直流伺服电机PID调节器的设计

针对数控火焰切割机自动调高系统的特点和要求,采用遗传算法对其位置调节器PID参数进行全局优化辨识,有效地解决了直流伺服电机的控制精度和快速性的问题.仿真结果表明,与传统的参数辨识法相比,遗传算法参数寻优使自动调高系统的动态性能大幅提高,具有很强的鲁棒性.     

面向对象规则和支持向量机的天宫一号高光谱影像分类

传统的高光谱分类方法通常基于单一像元的光谱或纹理特征,很少考虑地物空间结构信息与空间相关特征.本文将面向对象规则与基于像元的分类进行融合,利用对象的空间结构特征和光谱特征进行混合分类,旨在克服像元层次分类的不足.本文尝试性的提出了两种混合分类方法:(1)基于分形网络演化的多尺度分割支持向量机分类(2)基于多层分水岭分割的SVM分类,并将这两种方法应用到天宫一号高光谱数据上.结果表明:基于面向对象规则的混合分类方法有效地提高了分类精度,不仅能够改善同谱异物现象,而且解决分类结果中地物破碎的问题.     

导航卫星受地球外辐射带影响仿真分析

目前国际上不同国家已经进行了多个导航系统的开发,而构成这些系统的关键空间段实体导航卫星通常分布在中高轨道上,通过地球辐射带对该空间区域内卫星的辐射影响进行仿真分析,对于中轨道空间辐射,当卫星屏蔽厚度大于9mm,再增加屏蔽厚度的方法对辐射隔离不再敏感;对于高轨道空间辐射,当卫星屏蔽厚度大于7mm,再增加屏蔽厚度方法对隔离辐射已经不再敏感。可见,仅从辐射加固角度来说,在卫星设计时,以10年寿命为例,对于中轨道卫星屏蔽厚度区间取[8,9]mm为宜,对于高轨道卫星屏蔽厚度区间取[6,7]mm为宜。     

Impact of vegetation feedback on the mid-Pliocene warm climate

Studying the vegetation feedback during warm periods of the past can lead to better understanding of those in the future.In this study,we conducted several simulations to analyze vegetation feedback during the mid-Pliocene warm period.The results indicate that the main features of vegetation change in the mid-Pliocene were a northward shift of needleleaf tree,an expansion of broadleaf tree and shrub,and a northward expansion of grass,as compared to the pre-industrial period.The global annual mean warming ratio caused by vegetation feedback was 12.1％,and this warming ratio was much larger in northern middle and high latitudes.The warming caused by vegetation change was directly related to the surface albedo change and was further amplified by snow/sea ice-albedo feedback.     

Has climate change driven spatio-temporal changes of cropland in northern China since the 1970s?

Improving the understanding of cropland change and its driving factors is a current focus for policy decision-makers in China. The datasets of cropland and cropland changes from the 1970s to the 2000s were used to explore whether climate change has produced spatio-temporal changes to cropland in northern China since the 1970s. Two representative indicators of heat and water resources, which are important determinants of crop growth and productivity, were considered to track climate change, including active accumulated temperatures ≥10 °C (AAT10) and the standardized precipitation evapotranspiration index (SPEI). Our results showed that rapid cropland change has occurred in northern China since the 1970s, and the area of cropland reclamation (10.23 million ha) was much greater than that of abandoned cropland (2.94 million ha). In the 2000s, the area of cropland with AAT10 higher than 3,000 °C·d increased, while the area of cropland with an SPEI greater than 0.25 decreased compared to the 1970s, 1980s and 1990s. It appears that climate warming has provided thermal conditions that have aided rapid cropland reclamation in northern China since the 1970s, and drier climatic conditions did not become a limiting factor for cropland reclamation, especially from the 1990s to the 2000s. Approximately 70 % of cropland reclamation areas were located in warmer but drier regions from the 1990s to the 2000s, and approximately 40 % of cropland abandonment occurred in warmer and wetter conditions that were suitable for agriculture during the periods from the 1970s to the 1980s and the 1990s to the 2000s. Our results suggest that climate change can be considered a driving factor of cropland change in the past several decades in northern China, in addition to socioeconomic factors.     

多视角监控视频中动态目标的时空信息提取方法

监控视频中动态目标的精准定位与跟踪作为计算机视觉领域中重要的研究方向,近年来已成为监控领域的研究热点。传统视频动态目标检测仅依赖图像特征数据,忽略了与地理坐标系精准匹配,特别是对于多个摄像机覆盖的区域,拍摄的角度不同,投影后形成的图像空间分辨率也不同,因此,难以满足智能监控在复杂的地理场景中全方位时空信息感知。本文提出一种多摄像头协同的视频监控图像与地理空间数据互映射模型构建方法来获取动态目标的轮廓和地理位置等时空信息,首先建立监控图像信息与地理空间数据的互映射关系,将观测角度不同、尺度不同和空间分辨率的监控图像置于同一坐标系下,并在此基础上通过融合Canny算子与背景减法来检测目标的边缘信息;然后采用质心偏移算法还原目标在该场景的实际位置,从而实现多角度下连续跟踪,提升地理场景的时空理解力和分析力,提高动态目标的精准定位与跟踪能力。