基于小波的网络连接层流量模型

最近研究表明网络流量的统计特性在不同尺度上表现出很大的差异,由大量具有高斯性的网络流量和少量的突发性网络流量组成.为了更好地理解这种复杂的网络流量统计特性,引入多尺度分析工具小波,在多尺度分析的基础上提出一种新的基于小波的网络连接层流量模型,有机地结合了独立小波模型和多重分形小波模型的优点,即保持了独立小波模型在不同尺度对高斯性方差的拟合,同时保持了多重分形小波模型对突发性的拟合.真实网络流量的仿真结果表明这个模型能够灵活地适应网络流量在不同尺度下所表现出的高斯特性和突发行为.     

小波变换在大地回波噪声处理中的应用

基于小波变换的多尺度分辨特性,作者对连续信号、脉冲信号和宽平稳白噪声在小波域中不同分解尺度上的模极大值变化规律进行分析,根据它们的变化规律用Mallat快速算法设计出一套用于抑制低信噪比雷达回波中的干扰噪声、最大限度提纯真实回波信号的算法.仿真实验证明,该算法对抑制噪声作用明显,能较好地提纯雷达回波信号,这为低信噪比下雷达回波信号检测提供了一条有效途径.     

卫星干涉高光谱遥感图像序列差值压缩编码

根据干涉高光谱图像的成像特点和卫星遥感图像压缩编码系统要求，提出了一种新的卫星干涉高光谱图像序列压缩算法．新算法利用干涉成像光谱仪推扫成像特点提出了一种低存储量，帧间小波域匹配的图像序列压缩方案，提高图像质量3—4dB．算法首先通过小波域匹配算法检测出相邻图像之间的推扫位移量并移位求得两帧的最佳差值图像，然后对差值图像进行静止图像EBCOT压缩编码，从而提高总体编码效率．该算法具有与单帧图像编码算法相当的复杂度，只需要对当前图像与模板的差值图像进行基于小波变换的编码，避免了目前基于三维小波变换的编码算法对系统大存储量要求以及编码延时大的缺陷．仿真结果表明，本算法针对干涉高光谱图像编码，可获得比基于三维小波变换的算法相当甚至更好的效果，并且大大降低了编码延时．     

直升机机内PWM照明调光系统产生音频干扰问题的研究

在直升机机内照明调光领域，国外军机已广泛采用脉冲宽度调制（PWM）调光方式。PWM调光方式相比传统线性调光方式具有以下明显优点：效率高，调光范围宽，功耗小，带载能力强，附加损耗小等。但PWM调光系统的不足是电磁辐射，尤其会对机内通话系统产生音频干扰，在机通耳机中会出现明显干扰声，影响机内通话质量。本方案能有效抑制PWM...     

基于DSP和准同步算法的谐波监视控制器的研制

设计了一种多功能谐波监视控制器,该控制器基于DSP芯片TMS320C6713进行软硬件设计,采样运用准同步采样法.该控制器具有根据现场谐波污染状况进行滤波器投切功能,同时还可以通过监视器对现场移动物体侵入进行报警,具有良好的实际应用价值.     

面向叶片裂纹检测的自适应图像处理算法研究

提出了一种采用基于边缘检测的小波滤波和自适应阈值分割算法,实现叶片裂纹的自适应快速检测;仿真结果验证了这种检测算法的有效性,能满足外场对发动机叶片裂纹检测的要求。     

L׳ADROIT – A spaceborne ultraviolet laser system for space debris clearing

Small (1–10 cm) debris in low Earth orbit (LEO) are extremely dangerous, because they spread the breakup cascade. Pulsed laser active debris removal using laser ablation jets on target is the most cost-effective way to re-enter the small debris. No other solutions address the whole problem of large (~100 cm, 1 t) as well as small debris. Physical removal of small debris (by nets, tethers and so on) is uneconomical because of the energy cost of matching orbits. In this paper, we present a completely new proposal relative to our earlier work. This new approach uses rapid, head-on interaction in 10–40 s rather than 4 minutes, using 20–40 kW bursts of 100 ps, 355 nm UV pulses from a 1.5 m diameter aperture on a space-based station in LEO. The station employs “heat-capacity” laser mode with low duty cycle to create an adaptable, robust, dual-mode system which can lower or raise large derelict objects into less dangerous orbits, as well as clear out the small debris in a 400-km thick LEO band. Time-average laser optical power is less than 15 kW. The combination of short pulses and UV wavelength gives lower required fluence on target as well as higher momentum coupling coefficient. An orbiting system can have short range because of high interaction rate deriving from its velocity through the debris field. This leads to much smaller mirrors and lower average power than the ground-based systems we have considered previously. Our system also permits strong defense of specific assets. Analysis gives an estimated cost less than $1 k each to re-enter most small debris in a few months, and about 280 k$ each to raise or lower 1-ton objects by 40 km. We believe it can do this for 2000 such large objects in about four years. Laser ablation is one of the few interactions in nature that propel a distant object without any significant reaction on the source.     

三种时频分析在复合材料无损检测中的应用

随着复合材料的广泛应用和现代信号处理技术的发展,时频分析方法已广泛应用于各类复合材料的损伤检测。将短时傅立叶变换、小波变换和黄氏变换三种时频分析技术经过Matlab编程,结合Lamb波应用于玻璃纤维增强的复合材料层合板的无损检测,并对三种时频结果进行比较与分析。结果表明：上述三种时频分析方法都能进行复合材料的损伤判定,但黄氏变换方法的分析精度更高。     

石英半球谐振子精密成型技术

石英半球谐振陀螺以其轻质、低功耗、长寿命、高可靠、应用精度范围宽、成本低等优势，引起了惯性技术界的瞩目。半球谐振陀螺的核心器件——石英半球谐振子的成型技术是当前的研究重点。经不断试验和摸索，以范成球面弹性展成为基础，通过均衡磨削原则、磨头最大包络原则和三阶段37道工序的精细优化工艺流程，较好地实现了高精度微应力硬脆石英谐振子异形薄壁件的成型，形成了专用的结构成型中心，为石英半球谐振陀螺的核心半球谐振子球壳成型开创了一条可行途径。     

REMOTE SENSING IMAGE CODING METHOD COMBINING WAVELET TRANSFORM WITH CLASSIFIED VECTOR QUANTIZATION

A new remote sensing image coding scheme based on the wavelet transform and classified vector quantization (CVQ) is proposed. The original image is first decomposed into a hierarchy of 3 layers including 10 subimages by DWT. The lowest frequency subimage is compressed by scalar quantization and ADPCM. The high frequency subimages are compressed by CVQ to utilize the similarity among different resolutions while improving the edge quality and reducing computational complexity. The experimental results show that the proposed scheme has a better performance than JPEG, and a PSNR of reconstructed image is 31～33 dB with a rate of 0.2 bpp.