Highly efficient terahertz generation from periodically-poled RbTiOPO4

Terahertz (THz) generation by periodically-poled RbTiOPO4 (PPRTP) with a quasi-phase-matching scheme based on cascaded difference frequency generation (DFG) processes is theoretically analyzed. The cascaded Stokes and anti-Stokes interaction processes are investigated from coupled wave equations. The THz intensities and quantum conversion efficiency are calculated. Compared with that of non-cascaded DFG processes, the THz intensity in 7-order cascaded DFG processes is increased to 2.95 times. The quantum conversion efficiency of 149.9% in cascaded processes can be realized, which exceeds the Manley-Rowe limit.     

Theoretical analysis of stimulated polariton scattering from the A1-symmetry modes of KNbO3 crystal

Stimulated polariton scattering (SPS) based on noncollinear phase matching scheme from the A1-symmetry modes of KNbO3 crystal is investigated for generating terahertz (THz) wave. Frequency tuning characteristics of THz wave by varying the phase matching angle and pump wavelength are analyzed. The expression of the effective parametric gain length under the noncollinear phase matching condition is deduced. Parametric gain and absorption characteristics of THz wave in KNbO3 are theoretically simulated. The characteristics of KNbO3 for parametric oscillator (TPO) are compared with those of MgO:LiNbO3. The analysis results indicate that KNbO3 is an excellent optical crystal for TPO to enhance the output of THz wave.     

汽车发动机故障维修技术实践探究

汽车发动机作为汽车的核心部件,为汽车的运动提供最主要的动力来源,可以比作为汽车的"心脏",汽车发动机能否高效运转运转关乎着汽车的效能利用情况,汽车发动机使用寿命会受到部件设计、外界天气以及日常保养维护等因素的影响,这也是汽车发动机出现多种故障的主要缘由,所以做好发动机的故障检修以及日常保养是维护发动机正常运转的关键因素,并且发动机也是汽车所用组成构件中最容易出现各种故障现象的组件,目前常见的汽车发动机故障有四类,分别是发动机启动故障、发动机异常响声故障、发动机排气管放炮故障以及发动机的缺缸故障。本文就主要分析了这四种最为常见的汽车发动机故障,并就此提出相应的故障处理方法。     

基于机器视觉的疵点检测方法的研究进展

机器视觉在织物疵点检测方面具有广泛的应用,对纺织品质量控制具有重要的现实意义。本文通过对近几年相关文献的分析,简要综述关于疵点检测各种方法,并按照对织物图像处理的方法不同可分为:结构方法、统计学方法、频谱分析方法、基于模型的方法和基于学习的方法。着重阐述了各类方法的优缺点,并总结该领域的发展趋势。     

基于改进局部自适应对比法的织物疵点检测

为提高织物疵点检测的准确率和检测效果,采用了一种基于最相似邻域的背景估计法来进行检测。首先,利用同态滤波对图像进行预处理;然后,以滤波后图像每个像素点为中心点,以11像素×39像素的窗口大小为中心区域,通过计算中心区域与周围邻域的相似度,利用最相似的邻域进行背景估计;最后,利用背景差分原理获得目标图像,并采用阈值分割和形态学方法对图像进行处理,最终获得检测结果。实验结果表明:此方法优于传统的检测方法,不仅能够检测到复杂背景下的疵点图像,而且对不同组织及光照因素影响下的织物疵点图像同样具有很好的检测结果,检测准确率可达98%,具有较高的适用性与检出率,也具有一定的抗干扰性。