基于嵌入式技术的网络多媒体监控系统的实现

给出了一种基于嵌入式技术的网络多媒体监控系统 ,详细介绍了系统的方案和特点 ,网络摄像机的软硬件设计以及客户端软件功能与设计。     

一种适用于大规模变量的并行遗传算法研究

当前MapReduce并行编程模型得到了广泛的应用.相对于传统的基于PVM或者MPI的并行编程方式,它在执行时间和处理问题规模等方面有明显优势.针对并行遗传算法的特点,提出基于MapReduce实现一种典型的并行遗传算法——粗粒度并行算法的方法,用以解决大规模变量问题.实验平台采用Hadoop,硬件条件为普通的服务器集群.在多目标优化问题测试中,当问题规模达到一定、处理变量数超过10E+7时,并行算法效率比串行提高数倍,并且能突破内存瓶颈.根据MapReduce自身特点调整其参数,改变并行程度,分析其对并行执行时间的影响.     

Sensing of Damage Mechanisms in Fiber‐Reinforced Composites under Cyclic Loading using Carbon Nanotubes

The expanded use of advanced fiber‐reinforced composites in structural applications has brought attention to the need to monitor the health of these structures. It has been established that adding carbon nanotubes to fiber‐reinforced composites is a promising way to detect the formation of microscale damage. Because carbon nanotubes are three orders of magnitude smaller than traditional advanced fibers, it is possible for nanotubes to form an electrically conductive network in the polymer matrix surrounding the fibers. In this work, multi‐walled carbon nanotubes are dispersed into epoxy and infused into a glass‐fiber preform to form a network of in situ sensors. The resistance of the cross‐ply composite is measured in real‐time during incremental cyclic tensile loading tests to evaluate the damage evolution and failure mechanisms in the composite. Edge replication is conducted to evaluate the crack density after each cycle, and optical microscopy is utilized to study the crack mode and growth. The evolution of damage can be clearly identified through the damaged resistance parameter. Through analyzing the damaged resistance response curves with measurements of transverse crack density and strain, the transition between different failure modes can be identified. It is demonstrated that the integration of an electrically conducting network of carbon nanotubes in a glass fiber composite adds unique damage‐sensing functionality that can be utilized to track the nature and extent of microstructural damage in fiber composites.     

金川选厂3#系统磨矿分级多参数测控

介绍了金川选厂３＃系统磨矿分级多参数测控的原理。硬件组成和软件特点。该系统硬件配置经济合理，实用性强，控制软件及监控软件功能齐全，操作方便，稳定可靠，性能优越，投入运行后，提高砂石处理量５％，降低了生产成本，取得满意效果。     

基于活动轮廓的多分辨率自适应图像分割

本文在活动轮廓模型的基础上,提出了一种自适应图像分割方法,引入了新的图象统计信息、梯度信息有关的加权外部能量,使得分割结果与模型的初始位置无关,不受噪声影响;利用ACD方法使模型自适应地改变其拓扑结构;为了提高图象分的速度和鲁棒性,提出了多分辨率图象分割算法,利用该方法对一些形状、拓扑结构复杂的物体进行了分割实验,结果验证了该方法有效性。     

多进制小波变换及其在影像分析中的应用

首先简要地介绍了多进制小波基本理论并分析了其变换特征,然后阐述了其在影像缩小、边缘检测中的应用。     

γ能谱的数据获取系统

本文介绍了γ能谱测量实验室的数据获取系统,并介绍了为建立该系统而开发的一个通用IBM能信程序软件包COMLAN。该通信程序将建立在微型机上的数据获取系统与大型机(如PDP-11,VAX-11等)相连,一次最多可传送999个获取的谱数据到大型机上做离线处理,使微型机在大型机的后援下,用途更广途泛、使用更灵活。