SAMPLED-DATA STATE ESTIMATION FOR NEURAL NETWORKS WITH ADDITIVE TIME–VARYING DELAYS

In this paper, we consider the problem of delay-dependent stability for state estimation of neural networks with two additive time–varying delay components via sampleddata control. By constructing a suitable Lyapunov–Krasovskii functional with triple and four integral terms and by using Jensen's inequality, a new delay-dependent stability criterion is derived in terms of linear matrix inequalities(LMIs) to ensure the asymptotic stability of the equilibrium point of the considered neural networks. Instead of the continuous measurement,the sampled measurement is used to estimate the neuron states, and a sampled-data estimator is constructed. Due to the delay-dependent method, a significant source of conservativeness that could be further reduced lies in the calculation of the time-derivative of the Lyapunov functional. The relationship between the time-varying delay and its upper bound is taken into account when estimating the upper bound of the derivative of Lyapunov functional. As a result, some less conservative stability criteria are established for systems with two successive delay components. Finally, numerical example is given to show the superiority of proposed method.

History and evaluation of national-scale geochemical data sets for the United States

Six national-scale,or near national-scale,geochemical data sets for soils or stream sediments exist for the United States.The earliest of these,here termed the ＇Shacklette＇ data set,was generated by a U.S. Geological Survey（USGS） project conducted from 1961 to 1975.This project used soil collected from a depth of about 20 cm as the sampling medium at 1323 sites throughout the conterminous U.S.The National Uranium Resource Evaluation Hydrogeochemical and Stream Sediment Reconnaissance（NUREHSSR） Program of the U.S.Department of Energy was conducted from 1975 to 1984 and collected either stream sediments,lake sediments,or soils at more than 378,000 sites in both the conterminous U.S.and Alaska.The sampled area represented about 65%of the nation.The Natural Resources Conservation Service（NRCS）,from 1978 to 1982,collected samples from multiple soil horizons at sites within the major crop-growing regions of the conterminous U.S.This data set contains analyses of more than 3000 samples.The National Geochemical Survey,a USGS project conducted from 1997 to 2009,used a subset of the NURE-HSSR archival samples as its starting point and then collected primarily stream sediments, with occasional soils,in the parts of the U.S.not covered by the NURE-HSSR Program.This data set contains chemical analyses for more than 70,000 samples.The USGS,in collaboration with the Mexican Geological Survey and the Geological Survey of Canada,initiated soil sampling for the North American Soil Geochemical Landscapes Project in 2007.Sampling of three horizons or depths at more than 4800 sites in the U.S.was completed in 2010,and chemical analyses are currently ongoing.The NRCS initiated a project in the 1990s to analyze the various soil horizons from selected pedons throughout the U.S.This data set currently contains data from more than 1400 sites.This paper（1） discusses each data set in terms of its purpose,sample collection protocols,and analytical methods;and（2） evaluates each data set in terms of its appropriateness as a national-scale geochemical database and its usefulness for nationalscale geochemical mapping.

An Improvement on Data-Driven Pole Placement for State Feedback Control and Model Identification

The recently proposed data-driven pole placement method is able to make use of measurement data to simultaneously identify a state space model and derive pole placement state feedback gain. It can achieve this precisely for systems that are linear time-invariant and for which noiseless measurement datasets are available. However, for nonlinear systems, and/or when the only noisy measurement datasets available contain noise, this approach is unable to yield satisfactory results. In this study, we investigated the effect on data-driven pole placement performance of introducing a prefilter to reduce the noise present in datasets. Using numerical simulations of a self-balancing robot, we demonstrated the important role that prefiltering can play in reducing the interference caused by noise.

State Estimation of Distribution Network Considering Data Compatibility

Considering that the measurement devices of the distribution network are becoming more and more abundant, on the basis of the traditional Supervisory Control And Data Acquisition (SCADA) measurement system, Phasor measurement unit (PMU) devices are also gradually applied to the distribution network. So when estimating the state of the distribution network, the above two devices need to be used. However, because the data of different measurement systems are different, it is necessary to balance this difference so that the data of different systems can be compatible to achieve the purpose of effective utilization of the estimated power distribution state. To this end, this paper starts with three aspects of data accuracy of the two measurement systems, data time section and data refresh frequency to eliminate the differences between system data, and then considers the actual situation of the three-phase asymmetry of the distribution network. The three-phase state estimation equations are constructed by the branch current method, and finally the state estimation results are solved by the weighted least square method.

THE STATE SPACE RECONSTRUCTION TECHNOLOGY OF DIFFERENT KINDS OF CHAOTIC DATA OBTAINED FROM DYNAMICAL SYSTEM

Certain deterministic nonlinear systems may show chaotic behavior. We consider the motion of qualitative information and the practicalities of extracting a part from chaotic experimental data. Our approach based on a theorem of Takens draws on the ideas from the generalized theory of information known as singular system analysis. We illustrate this technique by numerical data from the chaotic region of the chaotic experimental data. The method of the singular-value decomposition is used to calculate the eigenvalues of embedding space matrix. The corresponding concrete algorithm to calculate eigenvectors and to obtain the basis of embedding vector space is proposed in this paper. The projection on the orthogonal basis generated by eigenvectors of timeseries data and concrete paradigm are also provided here. Meanwhile the state space reconstruction technology of different kinds of chaotic data obtained from dynamical system has also been discussed in detail.

State Estimation for Non-linear Sampled-Data Descriptor Systems:A Robust Extended Kalman Filtering Approach

This paper proposes a state estimation method for a class of norm bounded non linear sampled data descriptor systems using the Kalman filtering method. The descriptor model is firstly discretized to obtain a discrete time non singular one. Then a model of robust extended Kalman filter is proposed for the state estimation based on the discretized non linear non singular system. As parameters are introduced in for transforming descriptor systems into non singular ones there exist uncertainties in the state of the systems. To solve this problem an optimized upper bound is proposed so that the convergence of the estimation error co variance matrix is guaranteed in the paper. A simulating example is proposed to verify the validity of this method at last.

数据态档案管理模式概念、特征与挑战

在档案资源对象发展演化的过程中,数据态档案作为新型资源对象,具有信号和语义双重离散的本质特征。因此,有必要对已有档案管理理念和实践进行重塑,建立能够解除语义离散带来的高阶逻辑依赖的数据态档案管理模式。本文在档案资源连续空间演化图景的指导下,明确数据态档案管理模式的概念及功能定位,阐明其区别于其他形态档案资源管理模式的特征,并对模式构建所面临的主要挑战进行分析。研究认为数据态档案管理模式以体系性、连续性、可理解性和专业性为主要特征,在空间管控、证用价值维系、管理机制协同与技术能力建设等方面面临挑战。

程序调试中的数据结构演变可视化方法

数据结构是程序执行时的重要信息载体,其演变过程往往较为复杂和抽象。针对程序设计初学者较难理解数据结构演变过程的问题,提出一种面向程序调试中的数据结构演变可视化方法。在调试开始时,首先使用组件表达式表示数据结构,然后通过配置表以及布局函数将组件表达式生成可视化视图。在进行单步调试后,通过对比单步调试前后的可视化数据方式生成组件演变行为,运用有限状态机理论构建数据结构演变模型,该模型的执行使得组件发生演变,同时生成可视化动作。使用动画单元执行可视化动作,实现数据结构演变的可视化呈现,帮助初学者理解数据结构相关程序的执行过程。实验结果表明,将该方法应用于1个数据结构演变可视化框架StructV,并应用于1个在线可视化编程实训云平台的研发,可有效降低初学者理解数据结构演变过程的难度,并提升编程实训时的程序调试效率。

基于LabVIEW的离合器摩擦片性能试验台测控系统设计

针对摩擦片性能试验台架测量精度低、人工操作复杂的问题,开发了一种基于LabVIEW的摩擦片性能试验台测控系统,对系统关键信号采集并进行了信号处理,基于队列状态机设计了压力和转速的自动化调节方法。台架试验结果表明:该测控系统在监控采集和控制方面均满足实际需求,提高了试验台的自动化程度,具有一定的应用价值。

锂离子电池健康状态估计及寿命预测研究进展综述

随着锂离子电池的应用越来越广泛,锂电池健康状态的精确估计和剩余寿命的实时预测对于锂电池系统的安全运行和降低运维成本具有重要意义。锂电池内部复杂的物理化学反应和外部复杂工作条件,使得实现精准的健康状态估计和寿命预测具有挑战性。该文综述近年来锂电池健康状态估计和剩余使用寿命预测方法的研究现状,分析基于物理/数学模型、数据驱动、模型法和数据驱动融合,以及多种数据驱动融合的锂电池健康状态估计方法的优缺点及适用条件,并对比分析不同数据驱动类型的锂电池寿命预测方法。指出锂电池健康状态估计及寿命预测尚存在的问题,并对未来研究方向进行展望,对完善锂电池健康状态估计和寿命预测算法理论体系、指导实际应用技术具有重要意义。

烧结机煤气消耗量预测模型的性能比较研究

针对钢铁企业烧结机煤气消耗量预测精度较低的问题,研究建立差分自回归移动平均模型(ARIMA)、长短期记忆网络模型(LSTM)和极端梯度提升模型(XGBoost),用于预测烧结机的高炉煤气消耗量,利用钢铁联合企业的实际数据对比验证了预测模型性能。结果表明,XGBoost模型的预测精度高于ARIMA模型和LSTM模型。XGBoost模型的MAPE为3.45%,RMSE为703.53 m^(3)/min,R^(2)为99.91%,鲁棒性和泛化能力较强。此外,为了强化预测模型与烧结机不同运行状态间的联系,对烧结机不同运行状态的煤气消耗量进行预测。LSTM模型在烧结机正常生产状态表现出最好的预测效果,XGBoost模型则在烧结机减产和增产状态预测效果最佳。

非均衡数据下基于注意力网络和代价敏感学习的轨面状态识别

准确识别轨面状态,可为列车牵引/制动性能提升提供关键依据。重点针对传统代价敏感学习应用在非均衡轨面状态识别中存在的同类别样本重要性不同和多数类精度下降等问题,提出一种基于注意力网络和代价敏感学习的轨面状态识别方法。该法首先利用迁移学习思想将均衡数据集的特征迁移到非均衡轨面状态数据集,减轻少数类样本误分类影响;其次在骨干网络ResNet18中引入卷积注意力机制模块,增强网络对目标区域的特征学习能力和全局特征信息的感知性能,调整优化网络权重参数;最后构造依据轨面状态样本重要性大小的自适应加权平衡损失函数,降低决策边界对困难样本中多数类的过拟合,获得更加平滑的决策边界。非均衡数据下的实验结果表明,在3种非均衡比下,所提方法的准确率和召回率分别达到96.00%、90.67%、86.33%,与目前常用的方法Focal相比,分别提升了7.00%、2.34%、3.00%。此外,该方法在提高少数类召回率的同时可有效维持多数类的召回率,并且降低了网络训练时间成本。

基于数实结合的测控装备健康监测系统设计与实现

针对测控业务领域装备种类多、型号杂、布站分散、技术保障人力不足、远程技术协助手段落后等实际问题,结合测控装备自身特点和运维保障规律,提出了采用数字技术构建功能级数字测控装备,并利用数字装备和实体装备间物联网络使数实装备状态数据一致,实现基于数实结合的测控装备健康监测,实现装备状态数实同步,并可利用数实同步有效的评估装备状态、进行故障诊断和健康趋势预估,实现了运维保障数字化。

公立医院数据资产入表财务管理路径与框架研究

目的在遵循国家行政事业性国有资产管理相关法律法规和财政部《关于加强行政事业单位数据资产管理的通知》的前提下,探索公立医院数据资产财务管理的路径。方法按照《政府会计准则》和数据资产相关业务模式的特性,借鉴企业数据资产入表的路径,构建公立医院数据资产入表、会计核算、报告全流程财务管理框架。结果通过构建公立医院数据资产财务管理框架,切实保障了公立医院国有资产的安全与完整,规范了医院数据资产的财务管理。结论医院数据资产入表的财务管理路径与框架的创新构建,对后续出台相关政策、开展相关研究具有借鉴意义。

基于多源信息融合的船舶电气设备状态识别方法

为可靠掌握船舶电气设备状态,保证设备的运行安全,提出多源信息融合的船舶电气设备状态识别方法。采用时间序列模型检测并修正船舶电气设备多源历史数据中的连续异常数据和独立异常数据;基于联合卡尔曼滤波算法融合修正后的电气设备多源历史数据,依据融合后的多源数据训练谱聚类和深度神经网络,构建船舶电气设备状态识别网络模型,结合电气设备的实时运行数据,识别船舶电气设备状态。测试结果显示,该方法能够确定数据中的连续异常数据和独立异常数据,并且完成所有异常数据的修正,保证数据的完整性;离散度结果均在0.016以下;能够完成电气设备正常状态、异常状态以及紧急状态的识别,最小均方根误差值均在0.0044以下,识别效果良好。

基于状态可观测性和多模态数据PF的移动目标跟踪

为了实现对移动目标的跟踪,提出了一种新颖的基于目标状态可观测性和多模态数据粒子滤波(Particle Filter,PF)的跟踪方案。通过部署在目标移动区域中的传感器获得跟踪目标的距离和到达方向测量值,对接收到的数据进行预处理来计算PF的观测值,以形成一个临时距离图像。通过利用状态更新函数和形成的候选图像模板确定目标状态向量;在PF器中加入额外的加权阶段,使得PF器可自适应地同步多模态数据流,以实现鲁棒的目标跟踪。仿真实验结果验证了所提方案能够有效地跟踪移动目标。

状态依赖型切换系统的数据驱动方法建模

切换系统是由一系列连续或离散的子系统和切换机制组合而成的一类复杂系统,状态依赖型切换系统因其复杂性而尚未被深入研究.因此,通过系统的输入输出轨迹来对状态依赖的切换系统进行数据驱动建模,利用数据挖掘技术寻找数据之间的有用信息,建立输入与输出之间更形象的表达形式;在此基础上提出一种结构框架,根据辨识轨迹的切换时刻将数据分段,借助神经网络建立子系统的模型以及切换规则,深度挖掘状态依赖切换系统的信息,得到切换系统中子系统及子系统间的信息.实验结果表明:相比传统的机理建模,本文提出的数据驱动方法将建模的复杂度降低了17.3%.

基于扩张状态观测器的过热汽温系统建模与参数智能辨识

可再生能源的高渗透率给电网供需匹配带来严峻挑战的同时,燃煤机组需要承担着大量的调峰调频任务,这对过热汽温系统的安全稳定运行造成了一定威胁,因此有必要建立面向热工控制的汽温数学模型。考虑到迟延型扩张状态观测器(time-delayed extended state observer,TD-ESO)的总扰动信号中含有大量模型信息,提出一种基于ESO补偿模型的参数智能优化和信息提取方法,即以总扰动中未知信息量最小为目标,采用改进沙丘猫算法对模型参数优化并提取总扰动中已知模型信息补偿至ESO的输入端。在仿真算例方面,线性和非线性系统的测试结果表明,所提辨识方法对有无输入迟延的两种系统均有良好的适用性和较高的精度;在实际应用方面,基于超超临界二次再热机组的过热汽温系统数据进行模型辨识与验证,同样表明该建模方法是合理、准确的。因此,该文所建立的模型能够为汽温系统的控制策略设计和性能优化等方面提供有价值的参考。