野战有线通信系统二线全双工接口测试电路设计

野战有线通信系统全双工接口在单对线上同时收发信号,难以直接进行数据采集和测试。针对上述问题,研究了一种基于收发分离和自适应回波抵消的二线全双工接口测试方法。首先,设计电子收发分离器,实现线路上收发信号的隔离;然后,使用自适应回波抵消器,完成了近端回波干扰信号的抵消;最后,借鉴软件无线电技术和存储测试技术,构建测试系统,实现激励信号生成与数据采集。仿真和实验表明,该测试方法能够实现有效完成收发信号隔离以及回波干扰信号抵消,适用于野战有线通信系统二线全双工接口测试。     

基于DSP的音频会议信号合成算法研究

介绍数字音频合成和回波抵消的基本原理,提供了一种基于DSP的自适应回波抵消和归一化定标实现算法。仿真实验结果证实了该信号合成算法的合理性。实践中使用该算法能够达到很好的音频会议信号合成效果。     

基于DSP的回波抵消器的软件实现

随着通讯、数字信号处理和大规模集成电路技术的飞速发展,人们对语音通信质量的要求越来越高。回波抵消技术能有效的解决长距离电话网络、IP电话、免提电话和视频会议等通信系统中的回波问题,很好地改善了语音通信质量,具有广阔的市场前景。主要研究用于声音通讯系统中的声回波抵消技术。针对具体应用以及DSP芯片的特点,结合BEPNLMS算法,给出了一种基于TM S320VC 5402 DSP实现回波抵消的软件设计方案。     

用于回波抵消的最大长度序列相关近端语音检测算法研究

回波抵消是语音通信系统中不可缺少的一个重要组成部分。大部分回波抵消技术都是基于自适应LMS算法的。在实际应用中，近端语音检测的准确性会在很大程度上影响自适应LMS算法在双向通信环境下回波抵消的效果。本文提出了一种基于最大长度序列相关算法的近端语音检测算法。这种近端语音检测算法和自适应LMS算法相结合，得到的回波抵消算法在模拟双向通信环境下的回波抵消效果比自适应LMS算法高约8db。     

基于ITU-TG．165协议实现电话线路回波抵消的仿真研究

文章介绍了一种实现电话线路回波抵消的仿真算法。通过三个仿真单元的设计,最后实现了回波的抵消,实验结果表明,该算法能够成功地应用于系统中,取得了良好的效果。     

基于PN序列信道估计的回波抵消研究

针对估计序列影响信道估计精度并影响回波抵消效果的问题,从回波抵消基本原理出发,分析了PN序列作用信道估计的机理及其相关性对信道估计参数的影响,并与采用输入信号自身作估计序列的回波抵消方式进行了比较。对输入OFDM信号进行了仿真和实际运行。结果表明,采用PN序列估计信道,能有效提高估计的精度,改善回波抵消效果。     

基于并行滤波器的回波抵消技术方案

为了提高数字直放站回波抵消的收敛速度,首先研究了基于自适应滤波器的回波抵消技术,然后对其中的自适应滤波器的递推算法进行改进,形成了两个自适应滤波器并行计算、联合递推更新权值的技术方案。由于调节两个自适应滤波器权值的误差信号产生方式不同,方案可分为两种：方案一将回波抵消后的信号同时作为调节两个滤波器权值的误差信号(同时);方案二将天线接收信号与第一个滤波器输出信号的差值作为调节第一个滤波器权值的误差信号,而将该误差信号与第二个滤波器输出信号的差值作为调节第二个滤波器权值的误差信号(分别)。仿真结果表明,改进技术方案使回波抵消收敛速度提高11.11%~17.78%,从而有效改善了数字直放站回波抵消收敛速度慢的状况。     

基于声回波抵消两路算法的改进更新逻辑

声回波抵消两路算法被广泛用来检测系统双向通话;基于声回波抵消两路算法,提出了一种改进的控制更新逻辑。此更新逻辑通过比较滤波器的回波返回损失(ERLE),判断是否对滤波器进行更新。此改进更新逻辑能正确检测系统双向通话,避免滤波器的错误更新,并提高两路算法的收敛速度,减小存储器资源和计算量。仿真结果证实了此更新逻辑的有效性。     

声学回声抵消技术的研究

声学回声是影响视频会议系统和电话会议系统中通话质量的一种重要因素。本文研究了声学回声产生的原因和声学回声抵消器的原理及算法,简述了基于DSP的声学回声抵消器的实现方案,在设计中采用了NLMS算法,利用TI公司的DSPTMS320C5402作为处理单元实现了回声抵消器。     

基于盲信号分离的回波抵消技术

针对数字直放站回波抵消技术中自适应滤波法在多径回波信道条件下不能完全消除次径回波的问题,提出了基于盲信号分离的直放站回波抵消方法。首先对施主天线接收的混合信号进行相空间重构,使观测信号的数目大于等于独立信源的数目;然后利用独立分量分析法（ICA）对重构的信号进行盲信号分离;最后根据各分离信号和发送信号的相关情况判断有用信号,实现回波消除。对复杂多径回波信道条件下的多载波全球移动通信系统（GSM）信源进行回波抵消测试,分离得到的有用信号的相关系数可以达到0.9593。表明盲信号分离的方法可以实现复杂多径信道下的直放站回波抵消,有效解决了传统的自适应滤波法存在的问题。     

Preparation of dynamic posture and occurrence of low back pain.

Improper posture is considered as one of the causes for low back pain. This study focused attention on low back pain that occurs when people adopt a dynamic posture. Low back pain in attendants was investigated as a typical example of low back pain in a dynamic posture. When multi-dimensional quantification III was applied to the results of the investigation, low back pain was found to occur in six postural patterns: bending knee posture, the posture of lifting and holding a light object, the posture of tilting the trunk, working posture to push a cart, the posture of turning or lifting a cart, and the posture of stretching. This survey found that low back pain frequently occurred when an unexpected load was imposed on the lumbar region and experiments were conducted to simulate the unexpected loading of this region. The experimental results showed that lumbar muscular activity was not fast enough to cope with the load and resulted in an increased swaying of the trunk. This swaying was considered to induce a load on the lumbar region. It is concluded that low back pain can be prevented if an appropriate preparatory set is taken.     

Preparation of dynamic posture and occurrence of low back pain

Abstract

Improper posture is considered as one of the causes for low back pain. This study focused attention on low back pain that occurs when people adopt a dynamic posture. Low back pain in attendants was investigated as a typical example of low back pain in a dynamic posture. When multi-dimensional quantification III was applied to the results of the investigation, low back pain was found to occur in six postural patterns: bending knee posture, the posture of lifting and holding a light object, the posture of tilting the trunk, working posture to push a cart, the posture of turning or lifting a cart, and the posture of stretching. This survey found that low back pain frequently occurred when an unexpected load was imposed on the lumbar region and experiments were conducted to simulate the unexpected loading of this region. The experimental results showed that lumbar muscular activity was not fast enough to cope with the load and resulted in an increased swaying of the trunk. This swaying was considered to induce a load on the lumbar region. It is concluded that low back pain can be prevented if an appropriate preparatory set is taken.     

ZigBee技术在短程无线监控网络系统中的研究

随着无线网络技术和嵌入式系统的进一步融合,在日常生活中实现无线网络监控是现代社会发展的必然趋势。我们研究了当前几种主流的无线网络技术,通过比较和软硬件选择,我们认为具有低复杂度、低成本、低速率和低功耗特点的ZigBee技术比较适合低数据传输率短程无线监控网络控制系统的应用,而且我们在文中详细设计了一个可行的家庭无线监控网络系统的构建平台。     

基于低秩表示线性保持投影的特征提取算法

为了在特征提取过程中保持数据低秩特性不变,提出了一种基于低秩表示的线性保持投影算法用于维数约简。它能够使降维后的低维空间中的数据依旧较好地保持在原始高维空间中的低秩特性,准确地学习出数据的低维子空间。通过构建两个不同的低秩表示模型来 揭示两种不同结构特性的低秩权重,然后以保持数据的这两个低秩权重关系为目的来求解高维数据的低维空间。 在ORL库和Yale库人脸库上的实验结果证明,该算法比传统的特征提取方法更有效。     

改进的粒子群优化算法设计FIR低通数字滤波器

粒子群优化算法(PSO)因具有参数少、易于实现等优点,在解决优化问题时表现出很好的性能。有限长单位脉冲响应(FIR)数字滤波器因具有稳定的结构、易于实现等优点,在实际中有着很广泛的应用。因此,将基于三角函数因子的改进PSO算法(TFPSO)用于对FIR低通数字滤波器性能的优化,并将其与基于折射原理反向学习(refrPSO)、基于反向学习(OPSO)的PSO算法所设计的FIR低通数字滤波器的性能进行比较。在实验中构造出一种性能较好的适应值函数,以验证这几种改进的PSO算法所设计的FIR低通数字滤波器的性能。实验结果表明,基于三角函数因子的PSO算法滤波性能较差,而基于折射原理反向学习的PSO算法性能最佳。     

Using an adapted,task-level technology acceptance model to explain why instructors in higher education intend to use some learning management system tools more than others

Instructors in higher education perform some instructional tasks much more often using a learning management system (LMS) tool than other tasks. In studies that aim to explain these differences, the Technology Acceptance Model (TAM) perspective is missing. In this study, an adapted, task-level TAM questionnaire was used to measure task importance, task performance, LMS usefulness, LMS ease of use, and intention to use an LMS for 18 different instructional tasks among 180 instructors at a Dutch research university. The results show that low intention to use an LMS can be explained by (1) low task importance or performance, and/or (2) low LMS usefulness, and/or (3) low LMS ease of use level. The LMS tools and tasks within each of the three groups were not related substantively. This raises a question regarding whether an instructor's LMS intention level can best be explained by the combination of a specific tool, a specific instructional task, and a specific user interface.     

低碳背景下城市产业发展路径研究——以乌鲁木齐市为例

气候变化问题已经成为当今国际社会高度关注的全球性问题,发展低碳经济、建设低碳城市成为各国渐趋一致的看法。在低碳背景下,城市产业如何遴选,产业升级如何推进,这必将对传统产业发展路径的选择形成革命性的影响。本文以乌鲁木齐市为例,通过传统条件和低碳条件下城市产业发展路径的对比与选择,在传统工业行业中遴选出十大战略性产业予以重点发展,以推动乌鲁木齐市低碳产业发展,最终建设智慧城市、低碳城市,实现乌鲁木齐市的可持续发展和绿色发展。     

Alternative approach to low‐noise amplifier design for ultra‐wideband applications

Conventional ultra‐wideband low‐noise amplifiers require a flat gain over the entire 3.1–10.6 GHz bandwidth, which severely restraints the trade‐off spaces in low noise amplifier design. This article proposes a relaxed gain‐flatness requirement based on system level investigations. Considering the wireless transceiver front‐end with antenna and propagation channel, the unflat‐gain low‐noise amplifier with an incremental gain characteristic does not degrade the performance of overall system. As an alternative to its flat‐gain counterpart, the proposed unflat gain requirement tolerates gain ripple as large as 10 dB, which greatly eases the design challenges to low‐noise amplifier for ultra‐wideband wireless receivers. Two low‐noise amplifier examples are given to demonstrate the feasibility and design flexibility under the proposed gain‐flatness requirement. © 2007 Wiley Periodicals, Inc. Int J RF and Microwave CAE, 2007.     

Positive-mode dichroic LC display

There is considerable interest in large electronic display boards for presentation of information to mass audiences. Among the various display technologies, the dichroic dye liquid crystal display is considered best suited for such an application because ot its low power and low voltage requirements, viewability over wide angle and under high ambient lighting conditions, and potentially low fabricating cost.     

非线性联合学习的三维人脸表情合成方法

针对人脸表情数据的非线性分布特性,提出一种基于非线性联合学习的三维人脸表情合成方法.首先提出非线性联合学习理论,通过无监督回归将具有相同属性的三维人脸映射到相同的低维表达;其次,基于三维人脸的低维表达对低维表达进行重建操作,为给定的三维人脸合成表情,或基于样例表情进行表情的重定向.另外,非线性联合学习方法还能有效地处理带噪声及不完整的人脸数据,获得完整的表情人脸.实验结果表明,文中方法的表情重定向合成结果及合成效率优于已有方法.