转子故障的连续小波尺度谱特征提取新方法

引入图像分析方法,提出了直接从转子故障信号连续小波尺度谱中提取图像纹理特征的新方法.首先,通过转子故障模拟实验台采集了不平衡、不对中、碰摩及油膜涡动等典型故障信号;然后,分析了故障信号尺度谱的差别及所提取出的数字特征对故障的敏感性;最后用结构自适应集成神经网络进行了智能诊断实验,结果表明了本文所提出的尺度谱数字特征对转子故障诊断的有效性.      

基于概率网格Hough变换的多雷达航迹起始算法

 针对常规Hough变换(THT)航迹起始算法采用二元积累的弊端,提出了一种按测量点迹分布概率进行积累的多雷达航迹起始算法。首先把观测空间划分为统一大小的网格,然后根据雷达测量精度计算出观测数据落在每个网格内的概率,进而得到每个网格存在目标的概率,最后的Hough积累采用网格概率进行加权。与常规Hough变换法相比,新算法起始的航迹具有更高的精度和更好的鲁棒性,适合于不同精度雷达的联合航迹起始。仿真结果验证了算法的有效性。     

基于时-频分析的LFM信号检测研究

线性调频(LFM)波形是现代雷达系统广泛采用的一种非平稳扩谱信号波形,具有良好的低截获概率特性,为此,现代电子侦察系统常采用时频分析技术检测和分析这类时变信号.介绍了短时傅里叶变换和WVD等线性和非线性时频分析方法,运用这些方法对LFM信号的检测进行了分析,揭示了LFM信号在二维时频平面上的分布特征,并运用Hough变换检测WVD时频平面上呈直线分布的LFM信号.计算机仿真结果证实了此方法的有效性.     

稀疏泡状流中变形气泡的三维图像测量技术

研究气泡变形对于分析泡状流中气泡的受力和运动具有重要意义.设计了一种可确定稀疏泡状流中气泡空间坐标与变形参数的图像测量技术:原始图像经预处理和增强后,结合形态学方法识别出变形气泡投影的闭合轮廓.改进了分水岭方法对气泡投影进行分割,并用对称直线法确定投影中心改进Hough变换,得到椭球气泡模型的投影椭圆参数.提出了一种由两幅投影轮廓重构三维气泡模型的算法并分析了实验误差.该技术可较准确地测得气泡空间坐标和变形参数:投影椭圆的参数误差小于3个像素,气泡识别误差小于15%.     

大气潮汐的经典理论及有关现代模式的发展

按照历史发展, 回顾了从无耗无风背景下线性化的大气潮汐方程组出发得到一个二维定态变量可分离的基本偏微分方程及其全球Hough函数波模解的经典理论; 介绍了主要由Forbes和Hagan发展起来的在真实有耗有风背景下线性化的本质上仍是二维定态, 但纬度和高度变量不可分离的现代数值模式GSWM; 概述了近年来针对中高层大气的GCM (例如TIME-GCM, MUAM, CMAM和WACCM) 来研究潮汐非线性及非迁移潮汐等的发展趋势. 在融合理解的基础上对各模式的特点和模式间的差异进行了适当分析, 从中可以看出,半个世纪以来人们在从第一原理出发试图揭示全球大气潮汐的真实结构特征和长短期变化机制等方面进行了持续的努力, 这种努力又是与从利用地基雷达到卫星搭载仪器观测大气(特别是MLT区域的风场和温度场)的发展密不可分的.      

一种小波变换图象无失真编码方案

提出一种基于小波变换的图象数据无失真压缩编码方法,基本思路是利用小波变换实现图象的多分辨率分解,然后再对变换后的小波图象进行动态Ｈｕｆｆｍ ａｎ 编码,从而在不影响整个系统运行深度的前提下,最终实现压缩比较大的图象无失真压缩编码。     

空间大气成分探测傅立叶变换红外光谱仪

文章阐述了用于空间大气成分探测傅立叶变换红外光谱仪的国外发展状况，主要介绍了几种国外先进的高分辨率空间大气探测用傅立叶变换红外光谱仪的结构、应用范围和技术指标，比较了这几种设备的优缺点和技术特点。     

单片宽带频谱分析芯片设计

文章介绍了单片宽带实时频谱分析芯片的设计原理和方法,并在FPGA上进行了原型验证。设计中采用了基于多相滤波器组的下变频方案,并对下变频后每一路数据进行重叠加窗处理。为了减小芯片面积,对后端的FFT分析模块进行了优化设计,提出了将两条蝶形运算级合并为一条流水线的新型FFT设计构架。该设计可以不间断地处理数据,从而实现信号的无缝捕获。     

Analysis of precipitable water vapor from GPS measurements in Chengdu region: Distribution and evolution characteristics in autumn

The rainfall process of Chengdu region in autumn has obvious regional features. Especially, the night-time rain rate of this region in this season is very high in China. Studying the spatial distribution and temporal variation of regional atmospheric precipitable water vapor (PWV) is important for our understanding of water vapor related processes, such as rainfall, evaporation, convective activity, among others in this area. Since GPS detection technology has the unique characteristics, such as all-weather, high accuracy, high spatial and temporal resolution as well as low cost, tracking and monitoring techniques on water vapor has achieved rapid developments in recent years. With GPS–PWV data at 30-min interval gathered from six GPS observational stations in Chengdu region in two autumns (September 2007–December 2007 and September 2008–December 2008), it is revealed that negative correlations exist between seasonally averaged value of GPS–PWV as well as its variation amplitude and local terrain altitude. The variation of PWV in the upper atmosphere of this region results from the water vapor variation from surface to 850 hPa. With the help of Fast Fourier Transform (FFT), it is found that the autumn PWV in Chengdu region has a multi-scale feature, which includes a seasonal cycle, 22.5 days period (quasi-tri-weekly oscillation). The variation of the GPS–PWV is related to periodical change in the transmitting of the water vapor caused by zonal and meridional wind strengths’ change and to the East Asian monsoon system. According to seasonal variation characteristics, we concluded that the middle October is the critical turning point in PWV content. On a shorter time scale, the relationship between autumn PWV and ground meteorological elements was obtained using the composite analysis approach.