Robust H_∞ control of piecewise-linear chaotic systems with random data loss

This paper studies the problem of robust H∞ control of piecewise-linear chaotic systems with random data loss. The communication links between the plant and the controller are assumed to be imperfect （that is, data loss occurs intermittently, which appears typically in a network environment）. The data loss is modelled as a random process which obeys a Bernoulli distribution. In the face of random data loss, a piecewise controller is designed to robustly stabilize the networked system in the sense of mean square and also achieve a prescribed H∞ disturbance attenuation performance based on a piecewise-quadratic Lyapunov function. The required H∞ controllers can be designed by solving a set of linear matrix inequalities （LMIs）. Chua＇s system is provided to illustrate the usefulness and applicability of the developed theoretical results.

Model-based predictive controller design for a class of nonlinear networked systems with communication delays and data loss

This paper discusses the model-based predictive controller design of networked nonlinear systems with communica- tion delay and data loss. Based on the analysis of the closed-loop networked predictive control systems, the model-based networked predictive control strategy can compensate for communication delay and data loss in an active way. The designed model-based predictive controller can also guarantee the stability of the closed-loop networked system. The simulation re- suits demonstrate the feasibility and efficacy of the proposed model-based predictive controller design scheme.

Tracking Control of Wheeled Mobile Robots with Communication Delay and Data Loss

This paper considers the tracking control problem of a wheeled mobile robot under situation of communication delay and consecutive data packet dropouts in the feedback channel. A tracking controller in discrete-time domain for the case of ideal network condition is first derived, and then the networked predictive controller as well as two algorithms for dealing with communication delay and consecutive data packet dropouts are proposed. Simulation and experimental results verify the realizability and effectiveness of the proposed algorithms.

基于Focal Loss改进LightGBM的供水管网毛刺数据检测

针对数据不平衡导致的管网毛刺数据检测召回率偏低问题,提出一种Focal Loss改进LightGBM的管网毛刺数据检测方法。首先,结合管网毛刺数据的特点,针对性构造邻域相关特征。其次,将Focal Loss函数引入LightGBM,提高模型对难以检测的毛刺样本的权重,并对Focal Loss不同的参数取值进行实验,以平衡精确率与召回率。最后,选择不同参数的Focal Loss进行模型融合,进一步提升模型对不平衡毛刺数据的检测性能。在某市供水管网的真实数据上进行实验,结果表明,对比基于交叉熵损失函数的单一模型,本文提出的Focal Loss改进后的融合模型在毛刺数据上召回率和F1值的提升幅度达33.3和18个百分点,但毛刺数据的精确率还有待进一步提升。本文所提方法从损失函数入手,动态调整难易样本的权重,有效地提升了不平衡数据下的毛刺数据的检测性能。

THEORETICAL ANALYSIS OF IMPROVEMENT OF TRACK LOSS IN CLUTTER WITH MULTISENSOR DATA FUSION

The paper analyses the improvement of track loss in clutter with multisensor data fusion.By a determination of the transition probability density function for the fusion prediction error, one can study the mechanism of track loss analytically. With nearest-neighbor association algorithm. The paper we studies the fused tracking performance parameters, such as mean time to lose fused track and the cumulative probability of lost fused track versus the normalized clutter density, for track continuation and track initiation, respectively. A comparison of the results obtained with the case of a single sensor is presented. These results show that the fused tracks of multisensor reduce the possibility of track loss and improve the tracking performance. The analysis is of great importance for further understanding the action of data fusion.

A PMU data recovering method based on preferred selection strategy

Nowadays, the technology of renewable sources grid-connection and DC transmission has a rapid development. And phasor measurement units(PMUs) become more notable in power grids, due to the necessary of real time monitoring and close-loop control applications. However, the PMUs data quality issue affects applications based on PMUs a lot. This paper proposes a simple yet effective method for recovering PMU data. To simply the issue, two different scenarios of PMUs data loss are first defined. Then a key combination of preferred selection strategies is introduced. And the missing data is recovered by the function of spline interpolation. This method has been tested by artificial data and field data obtained from on-site PMUs. The results demonstrate that the proposed method recovers the missing PMU data quickly and accurately. And it is much better than other methods when missing data are massive and continuous. This paper also presents the interesting direction for future work.

A Brief Discussion of Data Storage Principle and Data Recovery Technology

According to relevant data statistics, each year, nearly 70% of users have disc data loss because of misuse, viral damage, physical damage and hardware failure, bringing irreparable damage to enterprises, institutions and individuals. So data recovery technology has attracted wide attention of users and how to use data recovery technology to help to recover lost data and to minimize the loss has become an urgent need. This paper starts from the software and hardware failure to analyze two aspects of hard disk data loss, data storage structure theory, and combined with practical experience, it elaborates data damage type and related recovery methods.

基于边界自适应SMOTE和Focal Loss函数改进LightGBM的信用风险预测模型

针对信用风险评估中数据集不平衡影响模型预测效果的问题,提出一种基于边界自适应合成少数类过采样方法(BA-SMOTE)和利用FocalLoss函数改进LightGBM损失函数的算法(FLLightGBM)相结合的信用风险预测模型。首先,在边界合成少数类过采样(Borderline-SMOTE)的基础上,引入自适应思想和新的插值方式,使每个处于边界的少数类样本生成不同数量的新样本,并且新样本的位置更靠近原少数类样本,以此来平衡数据集;其次,利用FocalLoss函数来改进LightGBM算法的损失函数,并以改进的算法训练新的数据集以得到最终结合BA-SMOTE方法和FLLightGBM算法建立的BA-SMOTE-FLLightGBM模型;最后,在LendingClub数据集上进行信用风险预测。实验结果表明,与其他不平衡分类算法RUSBoost、CUSBoost、KSMOTE-AdaBoost和AK-SMOTE-Catboost相比,所建立的模型在G-mean和AUC两个指标上都有明显的提升,提升了9.0%~31.3%和5.0%~14.1%。以上结果验证了所提出的模型在信用风险评估中具有更好的违约预测效果。

基于Focal Loss和卷积神经网络的入侵检测

入侵检测是信息安全防护领域中的一个重要环节。随着网络技术的发展,主动防御网络入侵变得越来越重要,同时入侵数据变得更加海量、复杂和不平衡,这导致传统的入侵检测技术的检测性能比较低,因此如何提高入侵检测系统的性能对于不平衡数据集的检测性能是一项巨大的挑战。传统的CNN模型对于处理复杂的数据具有很好的性能,但是在处理不平衡数据集上的效果不是很好。为了解决这个问题,提出一种基于Focal Loss和卷积神经网络的入侵检测方法。与传统的卷积神经网络不同,该模型利用Focal Loss损失函数来解决数据不平衡问题,并在卷积层加入正则化方法(DropBlock)用来提高模型的泛化能力。采用KDD 99数据集进行的实验表明,该模型入侵检测的准确率和精确率比传统的入侵检测模型有所提高。

Research on Data Fusion of Adaptive Weighted Multi-Source Sensor

Data fusion can effectively process multi-sensor information to obtain more accurate and reliable results than a single sensor.The data of water quality in the environment comes from different sensors,thus the data must be fused.In our research,self-adaptive weighted data fusion method is used to respectively integrate the data from the PH value,temperature,oxygen dissolved and NH3 concentration of water quality environment.Based on the fusion,the Grubbs method is used to detect the abnormal data so as to provide data support for estimation,prediction and early warning of the water quality.

Global parameter estimation of the Cochlodinium polykrikoides model using bioassay data

Cochlodinium polykrikoides is a notoriously harmful algal species that inflicts severe damage on the aquacultures of the coastal seas of Korea and Japan. Information on their expected movement tracks and boundaries of influence is very useful and important for the effective establishment of a reduction plan. In general, the information is supported by a red-tide（a.k.a algal bloom） model. The performance of the model is highly dependent on the accuracy of parameters, which are the coefficients of functions approximating the biological growth and loss patterns of the C. polykrikoides. These parameters have been estimated using the bioassay data composed of growth-limiting factor and net growth rate value pairs. In the case of the C. polykrikoides, the parameters are different from each other in accordance with the used data because the bioassay data are sufficient compared to the other algal species. The parameters estimated by one specific dataset can be viewed as locally-optimized because they are adjusted only by that dataset. In cases where the other one data set is used, the estimation error might be considerable. In this study, the parameters are estimated by all available data sets without the use of only one specific data set and thus can be considered globally optimized. The cost function for the optimization is defined as the integrated mean squared estimation error, i.e., the difference between the values of the experimental and estimated rates. Based on quantitative error analysis, the root-mean squared errors of the global parameters show smaller values, approximately 25%–50%, than the values of the local parameters. In addition, bias is removed completely in the case of the globally estimated parameters. The parameter sets can be used as the reference default values of a red-tide model because they are optimal and representative. However, additional tuning of the parameters using the in-situ monitoring data is highly required.As opposed to the bioassay data, it is necessary because the bioassay data have limitations in terms of the in-situ coastal conditions.

基于Datasheet的两电平逆变器损耗模型研究

为进一步提高市电领域中应用广泛的SPWM两电平逆变器输出效率并降低器件损耗,深入分析了双极性调制模式下的两电平无源逆变电源的工作过程,建立了基于数学分析方法的损耗模型并进行了仿真验证。采用基于非理想条件下的CMF10120D型SiC-MOS管与C4D05120A型SiC肖特基二极管(SBD),通过数据提取工具image2data与数据拟合方法获得Datasheet内功率管损耗与漏极电流、结温函数关系式,建立SiC-MOS与SBD的影响因子函数,最终得出两电平逆变器损耗模型。仿真分析表明所建立的模型正确可靠,脉冲宽度相对误差计算控制在0.75%以下,具有实用价值,能够为后续开展实验研究提供理论支撑。

PPDM中面向k-匿名的MI Loss评估模型

隐私保护数据挖掘(PPDM)利用匿名化等方法使数据所有者在不泄露隐私信息的前提下,安全发布在数据挖掘中有效可用的数据集。k-匿名算法作为PPDM研究使用最广泛的算法之一,具有计算开销低、数据形变小、能抵御链接攻击等优点,但是在一些k-匿名算法研究中使用的数据可用性评估模型的权重设置不合理,导致算法选择的最优匿名数据集在后续的分类问题中分类准确率较低。提出一种使用互信息计算权重的互信息损失(MI Loss)评估模型。互信息反映变量间的关联关系,MI Loss评估模型根据准标识符和标签之间的互信息计算权重,并通过Loss公式得到各个准标识符的信息损失,将加权后的准标识符信息损失的和作为数据集的信息损失,以弥补评估模型的缺陷。实验结果证明,运用MI Loss评估模型指导k-匿名算法能够明显降低匿名数据集在后续分类中的可用性丢失,相较于Loss模型和Entropy Loss模型,该模型分类准确率提升了0.73%~3.00%。

基于CNN-LSTM和卷复制方法的高可用系统设计方法

针对单机服务器存在的单点故障问题,以及主备双机中存在的逻辑故障导致数据丢失的问题,设计了一种基于卷积和长短期记忆神经网络(CNN-LSTM)和卷复制方法的HA(High Availability)系统。系统至少包含两个节点,一个主节点以及一个或多个备用节点,主节点和备节点之间支持主备切换。每个服务器节点上包含4个模块,分别是负责接收配置信息与读写请求的代理模块;进行磁盘读写操作和重定向读写的磁盘I/O(输入输出)模块;负责主备节点间备份快照、映射表、数据块复制的卷复制模块以及基于CNN-LSTM进行状态检测的高可用模块。实验表明,该系统不仅可以解决单点故障问题,也可以解决主备双机集群中无法解决的逻辑错误问题;同时基于CNN-LSTM方法,自动针对服务器的运行健康状态进行分析和预测,可以根据预测结果自动通知管理员进行处理或自动进行主备切换。

基于核心关联指标和虚拟等值电阻的台区理论线损计算

为提升传统线损计算方法的适用性和准确性,解决基于神经网络的线损计算方法的可解释性问题,提出了基于核心关联指标和虚拟等值电阻的台区理论线损计算模型。首先,提出了线损核心关联指标筛选方法,引用灰色关联度算法,计算线损关联指标与线损的相关性,筛选出核心线损关联指标,提高了模型的可解释性。其次,提出了物理-数据双驱动的理论线损计算模型,该模型考虑核心关联指标,以虚拟等值电阻替代传统等值电阻实现线损计算,提升了模型的适用性和准确性。然后,提出了基于二阶锥算法的参数估计方法,构建以理论线损改进值与实际值的方差最小的目标函数,以各类关联指标的临界值为约束条件,求解模型的参数。最后,应用实际监测数据和仿真模型数据,验证了文中所提方法的有效性和适用性。

基于手机信令HMM轨迹映射的交通方式识别算法

为解决手机信令数据稀疏性限制和重构轨迹特征提取与融合能力不足问题,提出一种出行轨迹重构与多源特征融合的交通方式精准识别算法。构建刻画基站信号传播路径损耗与信号强度的无线信号损耗模型,利用隐马尔可夫模型(hidden markov model,HMM)将手机信令轨迹由基站序列重构为路段节点序列,提出基于无线信号传播隐马尔可夫模型(wireless signal propagation hidden markov model,WP-HMM)的出行轨迹重构方法,用以描述信号强度与距离作用关系。基于出行重构后的轨迹,结合路段类型特征,提出了时空标准化相似性度量算法,以融合导航轨迹特征,并构建了基于随机森林(random forest,RF)的交通方式识别算法。实证分析表明:通过出行轨迹的重构,模型的平均识别精度提高了8%以上,且对新样本具有优异的泛化能力;相较于现有方法,时空标准化相似性度量算法能更准确捕捉轨迹间的移动模式;在不同环境下的轨迹识别中,模型在郊区区域的表现显著高于城区。所提算法在大规模手机信令数据的出行方式识别领域具有重要的应用价值。

基于熵变匹配追踪的叶端定时数据缺失识别方法研究

为解决叶端定时系统在实际应用中存在的数据缺失问题,提出基于熵变匹配追踪的叶端定时数据缺失识别方法。该方法利用相关熵诱导度量基于高斯核函数度量样本的权重。不同于正交匹配追踪对所有观测数据赋予相同权重,熵变匹配追踪基于相关熵诱导度量变化,对观测数据赋予不同范数类型的权重,使得其对异常值具有较好的鲁棒性。通过仿真分析与实验对该方法的性能进行验证,结果显示所采用的熵变权重因子为数据缺失位置分配了接近于零的权重,有效降低了数据缺失对特征提取结果的影响,证明了该方法的鲁棒性。基于熵变匹配追踪的叶端定时数据缺失识别方法为叶端定时系统的装机应用提供了理论支撑,具有技术借鉴价值。

基于多视数据重构损失标定的叶片型面光学检测

航发叶片型面检测是保证叶片加工质量、工作性能和使用寿命的关键环节。对此,该文开发一套基于线激光器的叶片型面检测平台;在完成线激光器与叶片相对位姿校准及Z向测量基准建立基础上,实现叶片型面特定截面处型线数据的多视采集;提出一种基于多视数据重构损失标定回转中心的型线重构方法。以两种典型燃气轮机导向叶片为检测对象,所提出方法获取的检测数据与三坐标测量机检测数据的对比结果表明,轮廓度检测平均绝对偏差在0.022 mm以内,最大标准偏差为0.0186 mm,最大RMS估计为0.0251 mm;特征参数检测最大偏差为0.0401mm。实验结果验证了所提出方法面向叶片型面检测的可行性。

基于知识蒸馏的不平衡数据下入侵检测方法研究

基于深度学习的网络入侵检测模型面临模型结构复杂、部署效率低及流量数据类别不平衡的问题.针对这些问题,提出了1种结合知识蒸馏和类别权重焦点损失的网络入侵检测方法.该方法以精度高、参数量较多的入侵检测模型作为教师模型,与小型学生模型生成蒸馏损失;引入增加类别权重的焦点损失函数作为学生损失;结合蒸馏损失与学生损失生成总的损失函数优化学生模型.实验结果表明,该方法性能相较于非蒸馏模型在各项指标上均有一定提升.

基于改进YOLOv5的变电站表计缺陷检测算法

准确检测变电站中的设备缺陷并及时进行处理是保证电力系统安全运行的重要措施.针对表计缺陷图像背景复杂、目标尺寸不一、外形差别大等问题,提出基于改进YOLOv5(you only look once的第5个版本)的变电站表计缺陷检测算法.为了提高泛化能力、解决训练过程中样本不平衡问题,利用旋转和改变图像亮度的方法进行数据增广.通过引入坐标注意力机制,在聚焦缺陷特征的同时,能突出缺陷特征的差异.为了使边界框回归更快速准确,将EDIOU loss(effective distance intersection over union loss)代替CIOU loos(complete intersection over union loss).实验结果表明:6种算法中,该文算法的准确度、召回率和mAP(mean average preciscion)均最高,分别达85.1%,86.6%,87.3%.因此,该文算法具有优越性.