基于结构支持向量机的目标检测跟踪同步算法

目标检测跟踪同步算法通过对视频帧的目标实时检测来达到跟踪的目的,该算法主要是为了维持一个能够在线训练的分类器,把从背景采样的样本作为负样本,从目标区域采样的样本作为正样本,然后通过分类器把二者区分开,以达到跟踪效果。然而当目标产生形变以及目标区域发生遮挡的时候,如何对样本采样和精确标记成为跟踪成败的关键。在结构支持向量机的框架下,提出一种基于结构支持向量机的目标检测跟踪同步算法。由于结构支持向量机的输出可以是复杂的数据结构,因此采用结构支持向量机,把目标位置估计作为结构支持向量机的输出,避免了对样本标记精确估计的需要,克服了当目标发生遮挡和大范围变形时导致的跟踪失败。仿真实验表明,该算法有良好稳定的跟踪效果。     

基于pinball损失的结构模糊孪生支持向量机

孪生支持向量机通过求解较小的二次规划问题,提高了分类器的性能,然而,该方法主要利用了类间可分的特性,并使用hinge损失函数构建相应的模型,它们并未充分考虑不同类中数据的结构信息以及不同样本对分类的影响,导致该方法对噪声具有较强的敏感性以及重取样的不稳定性.为了进一步提高孪生支持向量机的性能,基于pinball损失函数,将数据集中不同类的结构信息以及不同样本的作用引入到孪生支持向量机中,获得了基于pinball损失的结构模糊孪生支持向量机模型,从理论上导出了基于pinball损失的结构模糊孪生支持向量机算法pin-sftsvm,通过选取人工生成数据集与UCI标准数据集,对pin-sftsvm算法进行了实验,并与tbsvm、s-tsvm和pin-tsvm算法进行了性能比较,表明了提出算法的有效性.     

支持向量机的若干新进展

支持向量机是九十年代中期发展起来的机器学习技术,与传统的人工神经网络不同,前者基于结构风险最小化原理,后者基于经验风险最小化原理.实验表明,支持向量机不仅结构简单,而且技术性能尤其是泛化能力明显提高.本文是一篇综述,介绍支持向量机研究的一些新进展,希望引起大家的重视.     

基于支持向量机的C波段无线电异常信号识别

提出了一种基于支持向量机实现C波段无线电异常信号类型识别的方法.首先,通过对C波段实测异常信号进行统计分析,提取被识别信号的有效频域特征.其次,基于支持向量机分类器结构简单、泛化能力强、可获得全局最优等特点,构建了基于径向基核函数的支持向量机信号识别系统,取得了较高的识别率.最后,Matlab实验结果表明该方法信号识别效率高,在同等条件下优于神经网络.     

基于支持向量机和无源特征的目标识别方法

通过引入机栽雷达辐射这一无源特征.采用多类分类支持向量机进行类型识别,提出了一种更有效的目标识别方法.无源特征是有用信号和噪声的叠加.具有一定程度的不确定性,采用范数熵衡量无源特征,类间距较大,类内聚集性较强,还可以抑制噪声.支持向量机分类器结构简单、可获得全局最优、泛化能力强,多类分类支持向量机解决目标识别问题高效而且实用.实验证明,该方法明显地提高了目标识别的正确率.     

一种改进的支持向量机分类器

提出了一种将支持向量机分类和最近邻分类相结合的方法,形成了一种新的分类器.实验结果表明使用支持向量机结合最近邻分类的分类器分类比单独使用支持向量机分类具有更高的分类准确率,同时可以较好地解决应用支持向量机分类时核函数参数的选择问题.     

复杂分类问题支持向量机的简化

对于复杂分类问题,不可避免的会有错分情况,此时支持向量机的支持向量较多,影响了识别速度.为了解决这个问题,我们提出了基于最小错分间隔的分类思想,并在此基础上得出了一种新的简化支持向量机.与普通支持向量机相比,这种简化支持向量机有较少的支持向量、较高的识别速度,而且实验结果表明,它的识别精度完全可以与普通支持向量机的识别精度相媲美,甚至更优.     

支持向量机学习算法的研究现状与展望

回顾了支持向量机理论的发展历程,介绍了支持向量机的标准训练算法及其分解算法、变形算法、几何算法以及多类分类算法,重点描述了最小二乘支持向量机算法,总结了支持向量机理论及其应用的现状,对支持向量机的未来发展方向进行了展望。     

一种基于支持向量机的手写汉字识别方法

本文提出了一种新的基于支持向量机手写汉字识别方法.支持向量机作为一种新的机器学习方法,由于其建立在结构风险最小化准则之上,而不是仅仅使经验风险达到最小,从而使得支持向量分类器具有较好的推广能力.本文首先讨论了支持向量机的基本原理,然后,针对支持向量机识别大类别手写汉字所遇到的特殊问题,文章进行了分析和阐述,并在此基础上,提出了基于最小距离分类器预分类的两级分类策略.最后,针对GB2312-80的1034个汉字类别的120套手写样本,进行了实验仿真.实验结果表明,本文方法的汉字识别率较距离分类器有较大提高,其中多项式核函数的支持向量分类器,识别率平均提高3.38％,表明了本文方法的有效性.     

海域条件下大气折射率结构常数估算方法研究

针对海域条件下,激光通信大气折射率结构常数C2n估算方法存在的固有问题,提出了采用多项式拟合和支持向量机的方法进行C2n估算模型修正,从理论上改进了起伏方差和闪烁方差模型中结构参数依据经验的取值方式,进而搭建观测平台进行验证性实验.实验结果表明:在相同的大气环境条件下,采用多项式拟合和支持向量机方法进行C2n估算,其一致性要优于经典的估算模型,同时支持向量机方法较前者估算一致性优势明显.因此,本文提出的C2n估算方法可以为定量评价试验海域大气湍流程度提供理论依据.     

Heric逆变器驱动电路研究与设计

传统Heric逆变器各个桥臂采用相同的驱动电路,造成保护功能的重复,文章针对Heric逆变电路上下桥臂的导通状态,对逆变器上、下桥臂驱动电路的功能进行了划分,分别设计了上、下桥臂驱动电路,使其实现互补配合的保护功能,并针对驱动电路中光耦隔离芯片输出失真的问题,在光耦输入侧加入驱动增强电路。对驱动电路进行测试,实验结果表明该驱动电路可靠性较高,在占空比变化范围较宽时失真度小,逆变器输出波形较好。     

Fast amplitude control in twin-T bridge RC oscillators

When a sinusoidal voltage is applied to the symmetrical twin-T bridge, the voltages at the inner nodes of the bridge have equal amplitudes and are shifted by 90°. If they are squared and summed, the resulting DC voltage can be used to control the gate of a field-effect transistor connected to the bridge. The oscillator with such a control system simultaneously has low distortion and fast amplitude transient response.     

智能化数控调谐文氏电桥陷波器

介绍文氏电桥陷波器及其在失真度测量中的应用,提出一种智能化数控调谐的文氏电桥陷波电路.该电路利用AD536型真有效值检波器对文氏电桥陷波后的残余信号进行检测并进行模-数转换,由单片机控制文氏电桥的分档电容器和数控电位器,实现谐振频率的智能调谐.     

Voltage control in 3-phase stepped-wave inverters

Abstract

A number of voltage-control schemes in 3-phase bridge inverters are investigated. These control schemes are applied to dual 3-phase cascaded bridge inverters with fifth- and seventh-harmonic neutralization to realize pulse-width and amplitude controlled waveforms. kVA ratings of filter components are calculated to realize a total harmonic distortion of 5% in the output waveform. The control scheme with two pulses per half cycle spaced 60° apart in individual inverters gives the lowest total harmonic distortion and requires the lowest filter kVA rating. The theoretical results are experimentally verified and are in closo agreement.     

基于桥梁结构健康监测的无线传感系统自愈性研究

目前,对于大型桥梁结构健康监测的无线传感系统的应用研究正处于探索阶段,本文根据大型桥梁结构健康监测的需要,提出一种基于多跳的簇型无线传感器网络架构,以支持长距离、多测量点的大型桥梁结构健康监测。论文在介绍该系统的软硬件构架的基础上,重点阐述以备用中继节点方式实现网络自愈功能,使系统具有一定的自我修复功能,以提高系统的稳定性。     

A hybrid resonant converter operated as a low harmonic rectifierwith and without active control

A 1-φ high-frequency transformer-isolated AC-to-DC-controlled rectifier with low-line-current harmonic distortion using a variable-frequency-controlled hybrid parallel-series resonant converter bridge is proposed. Operation with and without active line-current waveshaping is presented. The state-space analysis method is used to get various design curves and to predict theoretically the waveforms and line-current harmonic distortion. By exercising active current control, power factor (PF) close to unity with low-line-current distortion is maintained for the entire range of operation in addition to zero-voltage switching (ZVS). Simulation results obtained from the state-space model, SPICE3 (without active control), and experimental results (rated at 150 W with and without active control) are given to demonstrate and validate the operating principle and characteristics of the proposed converter     

文氏电桥振荡电路设计方法

本文在详细分析典型文氏电桥正弦波振荡电路工作原理的基础上,推出一套该电路设计方法.笔者在阐述设计方法时,不仅给出了在波形指标要求下选择电路中元器件参数的计算公式和设计步骤,而且还介绍了抑制总谐波失真的方法与措施,并且通过一则设计实例及其仿真实验检验了该设计方法的正确性.     

Big data-enabled multiscale serviceability analysis for aging bridges☆

This work is dedicated to constructing a multi-scale structural health monitoring system to monitor and evaluate the serviceability of bridges based on the Hadoop Ecosystem (MS-SHM-Hadoop). By taking the advantages of the fault-tolerant distributed file system called the Hadoop Distributed File System (HDFS) and high-performance parallel data processing engine called MapReduce programming paradigm, MS-SHM-Hadoop features include high scalability and robustness in data ingestion, fusion, processing, retrieval, and analytics. MS-SHM-Hadoop is a multi-scale reliability analysis framework, which ranges from nationwide bridge-surveys, global structural integrity analysis, and structural component reliability analysis. This Nationwide bridge survey uses deep-learning techniques to evaluate the bridge serviceability according to real-time sensory data or archived bridge-related data such as traffic status, weather conditions and bridge structural configuration. The global structural integrity analysis of a targeted bridge is made by processing and analyzing the measured vibration signals incurred by external loads such as wind and traffic flow. Component-wise reliability analysis is also enabled by the deep learning technique, where the input data is derived from the measured structural load effects, hyper-spectral images, and moisture measurement of the structural components. As one of its major contributions, this work employs a Bayesian network to formulate the integral serviceability of a bridge according to its components serviceability and inter-component correlations. Here the inter-component correlations are jointly specified using a statistics-oriented machine learning method (e.g., association rule learning) or structural mechanics modeling and simulation.     

Six-freedom Displacement Measurement Based on Forward and Backward Laser Beam Paths

A displacement survey method that can be used in fields such as tunnel engineering, huge dam engineering and so on is introduced. According to this method, a series of measuring stations are set along a fold line, and a number of hi-direction laser beams and 2-D displacement sensors are used to realize the line and angle displacement measurements to obtain the distortion information of each point. The calculation formulas are given, and the errors are analyzed. This method can also be used in other distortion-measurement fields such as building, bridge and mountain distortion measurements.     

Boost bridge amplifier

A novel boost bridge amplifier which inherently doubles the power supply voltage, thus providing up to four times higher peak output power than a comparable state-of-the-art class-D amplifier, is proposed. The efficiency and the total harmonic distortion of both amplifiers during the amplification of a music signal are shown to be similar