Address pulses during sustain period for increasing luminous efficiency

Abstract— A new driving method using a short pulse applied to the address electrode (i.e., address pulse) during a sustain period is proposed to improve the luminous efficiency. In this method, short pulses are additionally applied to the address electrode during the rising or falling edge of the sustain pulses. In the case of a small sustain gap, address pulses synchronized with a rising edge of the sustain pulse can help the expansion of the discharge volume toward the address electrode, whereas address pulses synchronized with a falling edge of the sustain pulse produce a self‐erasing discharge, which improves the luminous efficiency. In the case of a large sustain gap, the application of the address pulse can produce a stable sustain discharge at a low sustain voltage level, generating an efficient discharge even in the case of the long discharge path.     

Classification of imprecise data using interval Fisher discriminator

In this paper, an imprecise data classification is considered using new version of Fisher discriminator, namely interval Fisher. In the conventional formulation of Fisher, elements of within‐class scatter matrix (related to covariance matrix between clusters) and between‐class scatter matrix (related to covariance matrix of centers of clusters) have single values; but in the interval Fisher, the elements of matrices are in the interval form and can vary in a range. The particle swarm optimization search method is used for solving a constrained optimization problem of the interval Fisher discriminator. Unlike conventional Fisher with one optimal hyperplane, interval Fisher gives two optimal hyperplanes thereupon three decision regions are obtained. Two classes with regard to imprecise scatter matrices are derived by decision making using these optimal hyperplanes. Also, fuzzy region lets us help in fuzzy decision over input test samples. Unlike a support vector classifier with two parallel hyperplanes, interval Fisher generally gives us two nonparallel hyperplanes. Experimental results show the suitability of this idea. © 2011 Wiley Periodicals, Inc.     

脉冲电源反应器的模糊控制研究

脉冲电源应用于污水处理时,需要反复调节交、直流耦合电路参数才能使反应器工作于流光放电模式。针对该方式存在的无有效可控量化指标及调节不准确问题,研究了反应器电压幅频特性以及反应器流光放电基本条件,提出了采用流光放电系数作为可控量化参数对脉冲电源反应器进行模糊控制,从而得到满足流光放电条件的交、直流叠加信号的方案。仿真及实验结果验证了该方案的可行性和有效性。     

多判别器循环生成对抗网络的素描人脸合成

素描人脸合成在娱乐和刑侦领域具有重要应用价值。为了解决传统素描人脸合成方法生成图像面部细节模糊,缺失真实感等问题,改进了CycleGAN网络结构,提出一种基于多判别器循环生成对抗网络的素描人脸合成方法。该方法选取残差网络作为生成网络模型,在生成器隐藏层中增加多个判别器,提高网络对生成图像细节特征的提取能力;并建立了重构误差约束映射关系,最小化生成图像与目标图像之间的距离。通过在CUHK和AR人脸数据库中的对比实验,证明了相比于原始CycleGAN框架该方法性能有明显提升;相比于目前领先的方法,所提方法生成的素描图像细节特征更清晰,真实感更强。     

振荡冲击电压下SF6气体中绝缘子气隙局部放电特性研究

随着GIS设备现场冲击耐压试验的推广实施,冲击电压下的局部放电检测被作为一种新方法,应用于GIS的现场绝缘诊断。冲击电压能够有效限制电晕稳定性作用,相比交流低频电压对局部极不均匀场类缺陷的发现能力更强,因此该方法将具有较好的应用前景。本文建立了冲击电压下的局部放电宽频带电脉冲检测系统,采用标准振荡雷电冲击(OLI)和标准振荡操作冲击(OSL)作为激励源,对绝缘子气隙缺陷在SF₆气体中的局部放电特性及影响因素进行了研究,并在此基础上针对预置人工绝缘气隙缺陷的实际GIS试品进行了模拟试验。研究表明：在振荡冲击下,气隙局部放电脉冲主要包括首次放电、后继放电及反向放电三种放电形式,其中首次放电对其他几种放电形式具有决定作用。且气压、背景场强及气隙尺寸比等对绝缘子气隙局部放电行为具有明显影响。GIS的模拟局部放电检测表明,两种振荡冲击电压(OLI和OSI)均能够有效检测GIS内部的气隙类缺陷。     

Extension of the gap statistics index to fuzzy clustering

The well-known gap statistic index proposed by Tibshirani et al. has successfully applied in many clustering evaluations. However, the gap statistic index cannot evaluate the clustering partitions from any fuzzy clustering algorithm. This is because fuzzy clustering cannot provide the within-cluster similarity measure that is used in the gas statistic index. Thus, the applicable range of the gap statistic index is very limited. In this paper, we present a new method that extends the gap statistic index to fuzzy clustering by using fuzzy membership notations. Our proposed method can extend the applicability of the gap statistic index, and outperform other existing fuzzy indices in several aspects. Experiments on eight sets of synthetic and real datasets are used to verify the applicability and efficiency of the proposed method.     

Equivalent circuit model and characteristic parameters for AC PDP discharge cells

In this paper, an equivalent circuit model for discharge cells of alternating current plasma display panel (AC PDP) is developed. In this model, a voltage controlled switch is introduced to describe the breakdown process of the gas discharge. And two zener diodes are introduced to describe the sustain process of the gas discharge. Three internal discharge characteristic parameters, the breakdown voltage of discharge gap, the minimum sustain voltage of discharge gap and the discharge cell capacitance ratio, are proposed. The formulas for expressing external discharge characteristic parameters by internal discharge characteristic parameters are derived. The experimental method for measuring internal characteristic parameters is developed and the measurement results for a macro discharge cell are presented.     

一种结合自注意和多尺度生成对抗网络的图像去雨方法

为去除雨天拍摄照片上的雨滴,针对被雨滴所覆盖区域未知,雨滴区域中大多数背景信息已经丢失,以及需要提升图像清晰度和对全局信息关注度的问题,在生成对抗网络中生成网络的自动编码器结构中添加自注意层,并在判别网络中引入多尺度判别器。通过注意力分布图的引导,自注意层的优化和多尺度判别器的评估,生成网络在关注雨滴区域的前提下进一步关注全局信息,多尺度判别器可由粗到细更好地判别雨滴图像与清晰图像之间的差距。实验完成了所提方法与其他方法的对比,以及自对比,并用峰值信噪比和结构相似性进行评估,结果表明了所提方法的有效性,其质量和指标数值均高于其他方法。     

基于定性仿真和模糊知识的离心式压缩机排气量不足原因诊断

针对引起离心式压缩机排气量不足的原因,本文提出一种基于定性仿真和模糊知识的诊断方法.利用压缩机结构原理和故障机理等定性知识建立故障定性模型库,并推理得到压缩机排气量受不同因素影响时的定性规则库.对系统变量的观测值利用定性趋势提取和模糊化进行定性化处理.提出了基于变量定性趋势和模糊定性值约束的滑动窗口加权匹配策略,并根据匹配结果诊断出导致排气量不足的原因.最后通过两个实例仿真验证了所提方法的有效性.     

一种新型的汽车智能雨刷控制系统设计

提出了一种基于模糊模型的自动雨刷控制系统设计方法。该系统通过红外雨量传感器检测雨量大小,使用模糊控制算法和脉宽调制技术调节雨刷器的速度;给出了系统的总体结构和具体设计方法。实验结果表明,本文所提出的自动雨刷控制系统性能可靠、反应灵敏、控制效果良好。     

Improved discharge time lag of address pulse in an ACPDP with auxiliary electrodes

Abstract— New driving waveforms are proposed for an ACPDP with an auxiliary electrode. Auxiliary pulses and a stepped scan pulse during the address period distinguish the proposed waveforms from conventional waveforms. The address discharge time lag in an ACPDP with auxiliary electrodes was improved by application of auxiliary pulses and a stepped scan pulse during the address period. The interaction between the auxiliary pulse and the stepped scan pulse generates priming particles directly prior to the address discharge, and these priming particles influence the address discharge. As a result, the firing voltage of the address pulse is lowered, and the minimum address voltage is lower than that of conventional driving waveforms. Experimental results confirm that the address discharge time lag of the proposed waveforms is 32% lower than that of conventional driving waveforms.     

基于条件能量对抗网络的肝脏和肝肿瘤分割

从图像中分割出肝脏和肝肿瘤是肝部疾病诊断重要手段之一,现有基于卷积神经网络（Convolutional Neural Network,CNN）方法通过为输入图像中每个像素分配类别标签来实现肝脏和肝肿瘤分割。CNN在对每个像素分类过程中没有使用邻域内其他像素类别信息,容易出现小目标漏检和目标边界分割模糊问题。针对这些问题,提出了条件能量对抗网络用于肝脏和肝肿瘤分割。该方法基于能量生成对抗网络（Energy-Based Generative Adversarial Network,EBGAN）和条件生成对抗网络（Conditional Generative Adversarial Network,CGAN）,使用一个基于CNN的分割网络作为生成器与一个自编码器作为判别器,通过将判别器作为一种损失函数来度量并提升分割结果与真实标注之间的相似度。在对抗训练过程中,判别器将生成器输出的分割结果作为输入并将原始图像作为条件约束,通过学习像素类别之间的高阶一致性提高分割精度,使用能量函数作为判别器避免了对抗网络训练中容易出现的梯度消失或梯度爆炸,更易于训练。在MICCAI 2017肝肿瘤分割（LiTS）挑战赛的数据集和3DIRCADb数据集上对提出的方法进行验证,实验结果表明,该方法不仅实现了肝脏与肝肿瘤的自动分割,还利用像素类别之间的高阶一致性提升了肿瘤和肝脏边界的分割精度,减少了小体积肿瘤的漏检。     

Fuzzy approach to select machining parameters in electrical discharge machining (EDM) and ultrasonic-assisted EDM processes

Condition monitoring of the machining process is very important in today's precision manufacturing, especially in the electrical discharge machining (EDM). This paper introduces a fuzzy-based algorithm for prediction of material removal rate (MRR), tool wear ratio (TWR), and surface roughness (R_z, R_k) in the EDM and ultrasonic-assisted EDM (US/EDM) processes. In this system, discharge current, pulse duration, and ultrasonic vibration of tool are the input variables and outputs are MRR, TWR, R_z, and R_k. The proposed fuzzy model in this study provides a more precise and easy selection of EDM and US/EDM input parameters, respectively for the required MRR, TWR, R_z, and R_k, which leads to better machining conditions and decreases the machining costs. The fuzzy modeling of EDM and US/EDM were able to predict the experimental results with accuracies more than 90%.     

基于车头时距交通信号的模糊控制系统的研究

提出了一种基于车头时距的绿灯时间修正算法的交通流模糊控制方案,该方案在自适应模糊控制算法的基础上,增加了绿灯时间修正算法模块,从而对绿信比进行了优化,并用Matlab软件进行了仿真研究。仿真结果表明,在相同的交通环境下,该方法比未进行绿信比优化的自适应模糊控制系统方案有效。     

基于转换生成网络的素描人脸识别

针对素描图像和可见光图像存在较大模态差异这一问题,提出一种基于转换生成网络的素描人脸识别方法,该网络同时实现跨模态图像生成和素描人脸识别.转换生成网络由生成器、判别器和特征转换网络(S网络)组成.生成器生成图像,判别器使得生成图像具备两种模态信息,S网络提取高级语义特征来辅助生成图像和识别.使用端对端训练来更新模型参数...     

Biased scan of plasma display panel for data voltage reduction

This paper proposes a method of reducing the data voltage V_d of plasma display panels (PDPs). The proposed biased-scan method uses two separate ground systems: one for the sustain pulse generator (FGND) and the other for the data address and control systems (CHGND). A dc voltage bias, which is applied between CHGND and FGND during the address period, reduces V_d while preventing the undesired glow discharge induced by a scan pulse only. CHGND is connected to FGND for the first sustain pulse of each subfield, which reduces the time lag of address discharge, but it is separated from FGND for the other sustain pulses to increase the margin of the sustain voltage. The proposed method was tested on a 15% Xe 50-in. Full HD (1920 × 1080) single-scan PDP which had a sustain discharge gap of 110 μm. V_d could be reduced by 20 V (30%), and the power consumption of the V_d voltage source decreased by ∼25 W (50%) from that of the conventional method.     

联合同步挤压小波变换和多尺度排列熵的局部放电类型识别

为稳定提取变压器局部放电信号的特征,提出一种基于同步挤压小波变换和多尺度排列熵的局部放电特征提取方法,再通过ＧＫ模糊聚类方法对局部放电信号的特征进行识别分类。首先,通过同步挤压小波变换对4种典型变压器故障产生的局部放电信号进行分解,将其分解为一组含有局部放电特征信息的模态分量;然后,通过多尺度排列熵量化各模态分量的局部放电特征信息,使用各模态分量多尺度排列熵的平均值作为识别特征向量;最后,利用模糊聚类得到的局部放电样本标准聚类中心,采用欧式贴近度进行局部放电识别分类。将提出的方法应用于变压器局部放电的实验数据上,并与基于小波分解方法和经验模态分解的识别方法进行对比分析,实验结果表明,所提出的方法具有更好的分类性,对变压器局部放电分类具有更高的识别精度,平均识别精度达到93.60%。     

基于T-S模糊神经网络的齿槽效应补偿方法研究*

针对齿槽效应带来的齿槽误差问题,提出在传感器探头内布设齿槽位置检测线圈,建立传感器齿槽特性模型和基于T-S模糊神经网络的齿槽补偿系统模型,依据齿槽位置信号对传感器进行齿槽误差补偿。利用附加动量的BP学习方法对网络进行学习和测试。仿真结果表明补偿模型的输出不再随齿槽位置波动,最大误差为依0.2mm,该种方法可以有效地消除齿槽效应并提高传感器的检测精度,满足高速磁浮车悬浮控制系统要求。     

电火花加工智能控制系统的建模及仿真

为增强电火花加工过程的稳定性,基于模糊RBF神经网络智能控制,同时利用模糊控制的解耦性,以在电火花加工过程中统计的空载率、正常放电率、不正常放电率(包括电弧及短路放电)、相对于脉冲周期和他抬刀周期的正常放电率为输入,设计了以伺服参考电压、脉冲间隔和抬刀周期为输出的电火花加工模糊RBF神经控制器,并利用MATLAB对设计的模糊RBF神经控制器进行了建模及仿真.仿真结果表明了模糊RBF神经网络控制系统在电火花加工中对于稳定加工状态的有效性.     

电厂锅炉液位监控系统

介绍了电厂锅炉液位与流量的监控方法和系统的特点,给出了它的硬件和软件,针对中小型电厂锅炉运行过程中的液位与流量要求,研究采用计算机与模糊控制技术,有效地保证了运行过程对液位与流量变化的要求。