基于Hessian矩阵的改进EDlines输电线识别算法

随着电力网络的飞速发展,采用无人机搭载高清摄像头来进行输电线巡检已成常态。为提高巡检的实时性和准确性,本研究提出了一种基于Hessian矩阵的改进EDlines输电线识别算法。首先,通过伽马变换对输电线图像进行预处理,利用Hessian矩阵特征值和特征向量以求取像素点主方向和主曲率,并获得输电线主体轮廓,从而摒弃了传统方法中梯度计算锚点和像素方向的繁琐步骤。接着,在主体轮廓基础上连接锚点以得到潜在直线线段像素链,并运用随机抽样一致性(RANSAC)算法来进行线段拟合。最后,根据直线间的距离和角度,迭代拟合以得到最终的输电线。实验结果表明,该方法能应对多种复杂环境下的输电线识别任务,抗干扰能力强,误检率显著降低,为高空输电线巡检提供了可靠的技术支持,具有重要的工程应用价值。

基于新型编解码网络的复杂背景航拍图像输电线识别

针对常规图像处理和现有语义分割方法从航拍图像中识别输电线速度慢、准确率低等问题,构建了新型高效的输电线识别编解码网络。为减少模型参数,提高计算效率,采用轻量级MobileNetV3作为编码器主干特征提取网络,并在浅层引出快捷链路与深层进行堆叠;通过金字塔池化模块(PSP)和深度可分离卷积构建解码器提高输电线多尺度特征复用能力并实现网络轻量化;采用跳跃连接结构级联编码器和解码器从而融合浅层和深层多尺度特征信息;利用迁移学习加快网络训练收敛速度并识别出输电线。实验结果表明,新型编解码网络能准确快速地识别出复杂背景下的输电线,MPA、MIOU和FPS分别达到了94.37%、86.95%和31 frame/s,识别精度和速度均优于UNet网络和PSPNet网络。

基于轻量型编解码网络的复杂输电线图像识别

针对现有输电线图像识别网络参数多、耗时长等问题,本文构建了轻量型编解码网络,实现了多根交叉复杂输电线的快速准确识别。编码器以常规MobileNetV3前16层为基础,通过减少网络参数,采用卷积块注意力模块代替常规MobileNetV3网络的挤压和激励注意力模块,从而提高了网络的输电线特征信息提取能力。结合深度可分离卷积和深度空洞空间金字塔池化模块构建解码器,扩大感受野,提高网络聚合不同尺度上下文信息能力。利用L1正则方法稀疏训练网络,根据缩放因子与对应通道输出乘积的数值,设定剪枝阈值去除网络冗余通道,有效压缩网络体积,提高输电线识别速度。实验结果表明,轻量型编解码网络的平均像素精度(MPA)、平均交并比(MIoU)和识别速度分别达到了92.11%、84.19%和41f/s,优于PSPNet、U2Net和已有改进的输电线识别网络。

全卷积网络电线识别方法

目的 电线预警对于直升机和无人飞行器的低空飞行安全至关重要,利用可见光和红外图像识别电线是一个有效途径.传统识别方法需要人工设计的滤波器提取电线的局部特征,再使用Hough变换等方法找出直线,支持向量机和随机森林等机器学习方法仅给出图像中有无电线的识别结果.本文提出一种基于全卷积网络的电线识别方法,能在自动学习特征提取器的同时得到电线的具体位置等信息.方法 首先利用复杂背景生成大量包含电线图像和像素标签的成对仿真数据;然后改进U-Net网络结构以适应电线识别任务,使用仿真数据进行网络训练.由于图像中电线所占的像素很少,因此采用聚焦损失函数以平衡大量负样本的影响.结果 在一个同时包含红外图像和可见光图像各4 000幅的电力巡线数据集上,与VGG(visual geometry group)16等多种特征的随机森林方法相比,本文方法的电线识别率达到了99%以上,而虚警率不到2%;同时,本文方法输出的像素分割结果中,电线基本都能被识别出来.结论 本文提出的全卷积网络电线识别方法能够提取电线的光学图像特征,而且与传统机器学习方法相比能将电线从场景中精确提取出来,使得识别结果更加有判断的依据.

基于OpenCV的电线颜色识别系统软件

随着计算机视觉技术的进步,工业上出现越来越多使用视觉对产品进行检测的现象,因此以电线为研究对象,经过图像的背景差分化、二值化处理,研究了提取电线轮廓的方法、提取电线特征向量的处理后,利用BP神经网络对提取到的颜色矩的特征数据进行训练,识别出电线颜色。经过实验,该系统对样本集的识别率为100%,表明该方法具有识别速度快、准确率高的特点。

基于车载激光扫描数据的电线杆自动识别与定位

本文对车载激光扫描数据进行了研究,将投影点密度(DoPP)与电线杆的几何特点相结合,提出了一种基于激光扫描数据的电线杆自动识别和定位的方法。首先,将车载测量系统获取的激光点云投影到水平面,建立水平格网,设置投影点密度阈值分类出杆状地物所在的格网区域;然后将分类得到的杆状数据拟合成为一条空间直线,以电线杆的几何特性作为期望值和方差,自动识别电线杆;最后,该空间直线与地面网格相交确定电线杆的定位位置。本文算法与T3D Analyst商用软件提取的电线杆进行比较,实验结果表明本文方法行之有效。

Reliability of flip-chip bonded RFID die using anisotropic conductive paste hybrid material

A reliability of flip-chip bonded die as a function of anisotropic conductive paste (ACP) hybrid materials, bonding conditions, and antenna pattern materials was investigated during the assembly of radio frequency identification(RFID) inlay. The optimization condition for flip-chip bonding was determined from the behavior of bonding strength. Under the optimized condition, the shear strength for the antenna printed with paste-type Ag ink was larger than that for Cu antenna. Furthermore, an identification distance was varied from the antenna materials. Comparing with the Ag antenna pattern, the as-bonded die on Cu antenna showed a larger distance of identification. However, the long-term reliability of inlay using the Cu antenna was decreased significantly as a function of aging time at room temperature because of the bended shape of Cu antenna formed during the flip-chip bonding process.

Some Key Technologies of a New-type CNC Curve Grinding Machine

This paper presented a new-type CNC curve grinding machine, in order to conquer the disadvantages of conventional NC curve grinding machine and improve surface quality, dimensional accuracy and machining efficiency. The new-type grinder adopts some high and new technologies: ① the normal tracing device could make grinding wheel coincide with the normal direction of the machining points in work piece, and improve surface quality and dimensional accuracy; ② the digital image on-line recognition system could monitor machining process and compensate wheel wear in real time; ③ linear motor drive mechanism could realize the exact adjustment of the grinding head digitally and increase the machining efficiency. The math models of normal tracing and circular tolerance zone were presented. The experimental results show that the new-type CNC curve grinding machine equipped with the three devices mentioned above is easy to realize the precision grinding of any complex curve, and improve surface quality, dimensional accuracy and machining efficiency.

Power swing and voltage collapse identification schemes for correct distance relay operation in power system

The new techniques were presented for preventing undesirable distance relay maloperation during voltage collapse and power swings in transmission grids. Initially, the work focused on the development of a fast detection of voltage collapse and a three-phase fault at transmission lines by using under impedance fault detector (UIFD) and support vector machine (SVM). Likewise, an intelligent approach was developed to discriminate a fault, stable swing and unstable swing, for correct distance relay operation by using the S-transform and the probabilistic neural network (PNN). To illustrate the effectiveness of the proposed techniques, simulations were carried out on the IEEE 39-bus test system using the PSS/E and MATLAB software.

ENHANCED ENERGY-SAVING ANTICOLLISION ALGORITHMS WITH IMPROVING THROUGHPUT FOR RFID SYSTEM

Tag collision algorithm is a key issue for energy saving and throughput with Radio Frequency IDentification (RFID) system more popular in sensing infrastructure of covering wider area on a large scale. Exploiting low energy consumption strategy would enable longer operational life of tags and reader with battery energy supply. And improving throughput is required on a large scale to preserve the capability of the correct reception. Therefore, this paper proposes an enhanced anti-collision algorithm called Dynamic Slotted with Muting (DSM), which uses multiple slots within a frame per node in a binary tree and takes tag estimation function to optimize the number of slots, and adds a mute command to put identified tags silence. The performance of the proposed algorithm is analytically provided, and simulation results show that DSM saves more than 40% energy consumptions both at reader and tags, and improves more than 35% throughput compared to the existing algorithms. Thus our algorithm is demonstrated to perform efficient energy savings at reader and tags with throughput improvement.

Application of PFMEA for Identification of Self-Recovering Failures in Production Lines for Automatic Assembly of Capacitors

Quality control is an important part of the capacitors automatic assembly process. Traditionally this control is being realized through a series of electrical measurements including capacity, tension, and tgct. In most cases, these measurements are not suitable for detecting defects that appear as a result of failures of certain elements of the assembly lines. These so called self-recovering failures very often remain unnoticed, because they do not cause a suspension of the assembly process and if not taken into consideration, they can seriously menace the final product quality. In this paper, the authors use PFMEA to identify and evaluate the risk of the self-recovering failures. They also propose a simple Simulink model, which could be useful when trying to estimate the effect of installing new control devices at an existing assembly line upon its overall reliability and productivity.

Design and simulation of self-powered radio frequency identification (RFID) tags for mobile temperature monitoring

This paper presents a design for a self-powered radio frequency identification (RFID) tag with a thin film bulk acoustic reso- nating piezoelectric power supply (PPS), which can be used for portable remote temperature monitoring. We call this system a PPS-RFID for short. The RFID systems have been found to have many applications in the internet of things (IOT) in the past decade. But semi-active RFID tags require an onboard battery which limits their applications in many fields. For these reasons, our research focuses on power sources for the RFID tags. This paper emphasizes the circuit design and simulation of PPS. In our tests, 0.283 mW was generated by PPS at 1 Hz vibration by a 650 N impact force. The results showed that the integrated PPS could supply sufficient power for the designed PPS-RFID tag. The PPS-RFID tag can be widely used for temperature monitoring during mobile transport of perishable items such as medicines or food.

A new image processing method for discriminating internal layers from radio echo sounding data of ice sheets via a combined robust principal component analysis and total variation approach

Discriminating internal layers by radio echo sounding is important in analyzing the thickness and ice deposits in the Antarctic ice sheet.The signal processing method of synthesis aperture radar(SAR)has been widely used for improving the signal to noise ratio(SNR)and discriminating internal layers by radio echo sounding data of ice sheets.This method is not efficient when we use edge detection operators to obtain accurate information of the layers,especially the ice-bed interface.This paper presents a new image processing method via a combined robust principal component analysis-total variation(RPCA-TV)approach for discriminating internal layers of ice sheets by radio echo sounding data.The RPCA-based method is adopted to project the high-dimensional observations to low-dimensional subspace structure to accelerate the operation of the TV-based method,which is used to discriminate the internal layers.The efficiency of the presented method has been tested on simulation data and the dataset of the Institute of Electronics,Chinese Academy of Sciences,collected during CHINARE 28.The results show that the new method is more efficient than the previous method in discriminating internal layers of ice sheets by radio echo sounding data.

Nonlinear identification of electro-magnetic force model

Conventional attractive magnetic force models (proportional to the coil current squared and inversely proportional to the gap squared) cannot simulate the nonlinear responses of magnetic bearings in the presence of electromagnetic losses,flux leakage or saturation of iron.In this paper,based on results from an experimental set-up designed to study magnetic force,a novel parametric model is presented in the form of a nonlinear polynomial with unknown coefficients.The parameters of the proposed model are identified using the weighted residual method.Validations of the model identified were performed by comparing the results in time and frequency domains.The results show a good correlation between experiments and numerical simulations.