容差条件下的模拟电路多故障诊断

针对容差模拟电路多故障诊断问题,提出了包括故障状态检测、故障粗糙集生成和故障元件定位的分步诊断方法。将有限测点的故障状态检测转化为容差约束下的线性规划问题;推导得到故障特征等式并生成故障粗糙集,证明了在容差条件下的故障特征等式中可选用标称参数电路的节点电压;对故障粗糙集中元件参数的偏差限值进行修正,对修正后的电路进行状态检测,实现故障元件定位,并估算故障参数偏差。仿真实例表明,该方法具有较高的诊断准确度和参数辨识精度。     

Identification technique of DC series arc‐fault strings in photovoltaic systems

The DC series arc‐faults lead a risk due to an aging of photovoltaic system components enhanced by environmental factors and high DC voltages in already existing photovoltaic systems. Until recently, many research activities have focused on the detection of DC series arc‐faults by current signature analysis. However, DC arc detector has two problems. First, it is difficult to distinguish arc noise and switching noise of power conditioner by circuit current in photovoltaic system. Secondly, when a DC series arc fault occurs in the photovoltaic system, DC series arc‐fault string cannot be specified because the arc noise is propagated in the circuit. In this article, we propose two novel techniques for the DC series arc‐fault detection and identification of faulty strings. Experiments were performed by generating DC series arc in the field. The proposed technique is investigated based on the experimental results in the terms of effectiveness and applicability.     

A generalized fault diagnosis method in dynamic analogue circuits

Fault diagnosis of analogue circuits is essential for analogue and mixed‐signal systems testing and maintenance. A new method is proposed in this paper for multiple fault diagnosis of linear analogue circuits in frequency domain. The Woodbury formula is applied to the modified nodal equation to construct the fault diagnosis equation, which relates the limited measured circuit responses with the multiple faults inside the circuit in a linear way. A recently developed ambiguity group locating technique is modified here to identify the faulty parameters directly. Computation cost is reduced compared to combinatorial search in traditional fault verification methods. Only one node is needed for voltage measurement, but multiple excitations on accessible nodes are required for fault identification. Parameter evaluation can provide the exact solution to the deviated values of faulty parameters. The faulty parameter deviations can have any finite values. Example circuits are provided to illustrate the proposed method. Two other methods for multiple analogue fault diagnosis sharing the same mechanism as the method proposed in this paper are also briefly described. The proposed method is extremely effective for the circuit with very limited accessible nodes and is also computationally efficient. Copyright © 2002 John Wiley & Sons, Ltd.     

Fault modeling on complex field using least‐square circle fitting for linear analog circuits

A new fault modeling method using least‐square circle fitting (LSCF ) for linear analog circuits is proposed in this paper. In this method, Monte Carlo simulation is used to obtain the output voltage values when the parameter of one faulty component is changed while those of the other components vary within their tolerance limits. All the output response voltage values for every faulty circuit statue are decomposed into real and imaginary parts on a complex field. Then, LSCF method is adopted to match these data, yielding a corresponding circular curve on complex plane, which also can be expressed with a circular function. Its center coordinates and radius are established as the fault features. During measurements, by calculating the distance from one real output to each circular center coordinate and comparing the distance with each circular radius, a faulty component can be diagnosed. The effectiveness of the proposed approach is verified by experiments. The results show that (i) the proposed fault modeling can accurately locate the fault component, and (ii) it can also simply be a fault dictionary. © 2017 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于波形相似度的容差模拟电路软故障诊断

针对容差模拟电路输出波形参数的随机性,构造了描述特定电路状态的特征区间向量,据此提出了一种基于多频响应波形相似度的软故障诊断方法。对每一种软故障,分别施加K种不同频率的激励信号,获取K个由输出响应构成的标准特征区间向量作为该故障状态的信号特征。对待测电路施加相同的K种频率的激励信号,获取K个待测样本,分别与每种故障模式的K个相应的标准特征区间向量进行相似度计算,获取K个波形相似度,在波形相似度融合中考虑了均方差权重系数和可信度权重系数,得到待测样本与该故障模式的总体相似度。最后依据一定的模糊规则进行故障定位。仿真实验证明,本方法在保证较高诊断准确率的情况下,对测点要求低,易于在自动测试系统中实现诊断的自动化。     

Diagnosis of Short‐Circuit Faults in Stator Winding Inside Low‐Voltage Induction Motor Using Impulse Voltage Test

Short‐circuit faults in windings due to the deterioration of insulation is among the most common faults in motor drive systems. An easy and effective fault diagnosis method is urgently required to ensure highly reliable operation. This paper proposes a novel method for the diagnosis of short‐circuit faults in stator winding inside a low‐voltage induction motor without removing the rotor, by performing an impulse voltage test. As the rotor does not need to be removed from the motor in this novel diagnosis method, the method can be put to practical use. In this study, first, several impulse voltage tests are carried out on the stator windings of motors. Second, the values of two features that represent the characteristics corresponding to the condition of the motor are calculated, and it is found that the shape of the feature distribution does not depend on the rotor position but on the condition of the winding. Third, the distance between the feature distribution for the healthy motor and features obtained from a target motor is calculated. On the basis of this distance, the condition of the stator winding inside the induction motor is determined. The effectiveness of the proposed diagnosis method is verified by performing experiments that involve several motors with healthy and faulty windings.     

基于多目标线性规划的模拟电路软故障诊断方法

阐述了一种新的容差条件下基于多目标线性规划理论的模拟电路软故障诊断方法。通过灵敏度分析建立电路测试节点电压增量方程,并以元件参数变化量与标称值的百分比作为故障判据。采用多目标线性规划方法求解测试节点方程组来估算电路中各元件在一定故障情况下的参数变化百分比,将方程组的解与各元件容差范围相比较来定位故障元件。仿真结果表明,该方法兼顾故障元件的定位和故障元件参数变化量的估算,可以有效地实现模拟电路元件参数在一定容差范围内变化条件下的软故障诊断。     

Non‐linear analog circuit fault diagnosis with large change sensitivity

Large change sensitivity has been proved efficient at, but restricted to, generating a linear circuit fault dictionary. This paper discusses the extension of large change sensitivity to non‐linear analog circuit fault diagnosis. The fault dictionary is divided into d.c. and a.c. sections. In the d.c. domain, non‐linear components are approximated with piecewise linear models. By relating the operating region of each piecewise linear model to the magnitude of a single fault in a procedure termed preconditioning, it is shown that large change sensitivity can efficiently compute the response of a faulty non‐linear circuit. Results presented of an analysis of computational complexity show a significant reduction in the cost of simulating single linear resistor faults in a non‐linear circuit using this method. In addition, after establishing that the resistive portion of the circuit is fault free, a fault dictionary is constructed for dynamic components using large change sensitivity in the small signal a.c. domain. Included with a discussion on the issues of large change sensitivity based simulation‐before‐test, a small non‐linear circuit is used to demonstrate the effectiveness of the proposed fault diagnosis algorithm. Copyright © 2000 John Wiley & Sons, Ltd.     

非线性电路多软故障的智能优化递阶特征选择诊断方法*

针对非线性模拟电路的多软故障诊断中由于部分状态特征近似而影响诊断准确率的问题,在不提高电路信息采集工作量的情况下,采用分层诊断的思路,提出了基于Wiener核的智能优化递阶特征选择方法。该方法在获得电路各种状态的Wiener核后,采用智能优化算法对各状态的核特征进行特征选择,以代表各状态的特征构成的矢量的集总欧氏距离为评价函数,对集总欧氏距离的最大值进行寻优得到最优解;再对各个特征矢量间的距离进行判别,找到相互距离小于设定阈值的各个状态,构成次阶故障状态类,对该类故障状态采用前述的方法进行智能优化故障特征选择,得到次阶各状态的最优特征矢量;以此类推,直到得到满意的分辨率为止。实验表明,该方法有效地提高了多软故障诊断的准确率。     

模拟电路统一软故障诊断的研究

在模拟电路故障诊断的诸多方法中,模拟电路统一软故障诊断方法成为新的研究热点。从模拟电路统一软故障诊断方法的提出,到模拟电路统一软故障诊断方法的改进,将其与基于多种方法融合的模拟电路软故障诊断方法做了比较。最后重点分析了模拟电路统一软故障诊断的研究现状和存在的不足,基于实用化的角度,从统一的模拟电路故障诊断的功能模块划分原则和方法、通用的模拟电路故障可诊断性分析方法、通用的模拟电路模块化分级诊断软故障诊断方法、统一的诊断效果评估指标4个方面指出了模拟电路统一软故障诊断的发展趋势。     

Signal analysis‐based fault classification and estimation of fault location of an unbalanced network using S‐transform and neural network

This paper demonstrates a technique for the diagnosis of the type of fault and the faulty phase on an overhead transmission line, followed by locating the particular fault on the affected phase. The power system network considered in this study is a three‐phase transmission line with unbalanced loading simulated in the PowerSim Toolbox of MATLAB. S‐transform is used to compute the energy components of the voltage signals of the three phases of the transmission line. These features are used as input vectors of a probabilistic neural network (PNN) for fault detection and classification. Detection of the faulty phase(s) is followed by estimation of fault location. The voltage signal of the affected phase is processed to generate the S‐matrix. The frequency components of the S‐matrices for different fault locations are used as input vectors for training a backpropagation neural network (BPNN). The results are obtained with satisfactory accuracy and speed. All the simulations have been done in MATLAB environment for different values of fault locations, fault resistances, and fault inception angles. The effect of noise on the simulated voltage signals has been investigated. The analysis has been further extended by implementing the proposed method in a modified version of IEEJ West 10 machine system model. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Frequency‐domain grouping robust fault diagnosis for analog circuits with uncertainties

A new frequency‐domain grouping robust fault diagnosis (GRFD) scheme, based on the design for both grouping‐robust estimation/evaluation of component variation rate and Boolean‐based decision process, is proposed for solving fault diagnosis of large‐scale analog circuits with uncertainties, including both component tolerances and measurement errors.In this scheme, first, grouping‐robust estimators for one partitioning of all component groups, designed by using grouping‐full‐column rank output measurement technique, are employed to estimate the component variation rate. Secondly, based on the faulty possibility judgment policy suggested in this research, grouping‐robust evaluators are employed for the careful evaluation of component variation rate in order to reduce the effect of both component tolerances and measurement errors on the diagnostic test result for one partitioning. Next, both repeating the same one‐partitioning GRFD for a variety of partitionings and combining Boolean‐based decision process, an overall GRFD scheme is proposed for achieving analog fault diagnosis. Finally, a fault diagnosis example of a large‐scale analog circuit is illustrated. The results indicate that without any a priori information about the type of component failure, a precise t‐diagnosable case for a large‐scale circuit with component tolerances and measurement errors is accomplished at the lowest measurement cost. Copyright © 2001 John Wiley & Sons, Ltd.     

基于改进CEEMD和RF的低压串联故障电弧识别方法

了解决完整集合经验模态分解(complete ensemble empirical mode decomposition, CEEMD)得到的固有模态函数分量数目及其频段不固定,以及故障电弧特征难以准确提取导致故障识别准确率低的不足,引入T检验和方差贡献率形成了一种改进CEEMD方法,进一步提出一种基于改进CEEMD和随机森林(random forest, RF)的串联故障电弧识别方法。首先,依托串联电弧故障试验平台,采集不同负载的电流信号。然后,采用改进CEEMD对信号进行分析并提取故障特征量,以TreeBagger函数进行特征降维,形成特征向量样本集。最后,结合RF构建故障电弧诊断模型,对样本集进行分类识别。结果表明：改进CEEMD能有效地提取不同负载电流的故障特征,所提故障电弧识别方法的识别准确率达到97.50%。通过进行不同特征提取方法和不同分类模型对诊断结果影响的消融实验,进一步证明了所提方法的可行性。     

一种基于径向基神经网络和频谱分析的电路元件故障诊断方法

针对一类模型不确定的非线性系统,提出了具有强鲁棒性和高灵敏度的在线故障检测与诊断方法。其中,系统只有输入、输出可检测,故障是关于输入和状态的非线性函数。将RBF神经网络和频谱分析相结合,由RBF神经网络来学习及存储电子电路的故障频谱和故障类型之间的映射关系,介绍了该算法的实现过程。并以某船舶电气设备放大电路为例建立仿真系统。仿真结果和实验实例表明,该算法可以快速有效地对故障元件进行定位,识别率较高。     

基于小波变换和CFA-LSSVM模拟电路故障诊断

为了提高模拟电路软故障诊断、识别的正确分类率,提出了一种提升小波变换和混沌萤火虫算法(CFA)优化LSSVM参数的模拟电路故障诊断方法。首先对采集到的被测电路输出电压信号进行提升小波变换;然后对变换后的数据进行因子分析法对优化处理,将经优化的数据作为不同模式的故障特征集;最后将所得故障特征集作为样本输入到CFA-LSSVM模型进行故障诊断。实验结果表明,该方法的故障诊断正确率达到了98%以上,提高了诊断性能,可适用于模拟电路的故障诊断。     

Concurrent analogue fault simulation,the equation formulation aspect

Fault simulation is an essential tool for developing test patterns for circuits. Because the potential number of faults in a circuit is potentially very large, computational efficiency is an important consideration. In the digital domain, concurrent fault simulation is well‐established as an efficient tool. For analogue circuits, fault simulation is often performed by repeated insertion of possible faults and resimulation of the circuit. Consequently, methods for efficient concurrent analogue fault simulation are attracting attention. A review of existing methods of concurrent analogue fault simulation shows that most are based on a similar fundamental perturbation of the original fault‐free circuit equations, although the methods differ in the procedure applied after the circuit equations are formulated. We develop here a comprehensive set of element stamps, describing faulty elements, enabling effective and routine equation formulation for faulty circuits. These may be used no matter what method of fault simulation is later applied. These stamps are used in a new technique for concurrent analogue fault simulation, based on modified nodal analysis. A significant improvement in efficiency, compared with other methods, is demonstrated. Copyright © 2004 John Wiley & Sons, Ltd.     

基于强跟踪滤波器的模拟电路故障在线诊断方法

基于强跟踪滤波理论,给出了模拟电路故障实时诊断的一种新方法.该方法以建立模拟电路状态空间模型为基础,利用强跟踪滤波器对电路状态及元件参数进行估计,当元件参数发生软、硬型故障时,根据强跟踪滤波器对元件参数的跟踪结果及修正的Bayes分类算法,可实时诊断模拟电路中的元件故障.此外,该方法对于元件容差引起的故障诊断模型的不确定性具有较强的鲁棒性,对非线性电路的故障诊断也非常适用.实验结果证明了该方法的有效性.     

基于改进CEEMDAN分解与时空特征的低压供电线路串联故障电弧检测

针对低压线路中的串联故障电弧检测难题,提出基于改进CEEMDAN分解与时空特征的串联故障电弧检测方法。首先,采用CEEMDAN算法实现电流信号的完备分解,并以各IMF分量的峭度指标、裕度指标、能量特征和能量熵特征为判定依据,实现高频段信号粗选。然后,提出空间尺度和时间尺度相融合的特征构建方法,捕获各粗选高频IMF分量的局部特征,增强电流特征对比度和判别力。最后,采用子空间变换算法实现电流时空特征集合的二次降维,并基于SVM实现串联故障电弧检测。实际试验证明,所提算法的平均故障电弧检测准确率达88.33%,能够实现高效的串联故障电弧检测。     

基于灵敏度特性函数的特征提取与故障诊断

为解决电路故障诊断时故障可靠分类以及特征信息有效提取的问题,提出了一种基于灵敏度特性的故障样本分类和故障特征信息提取方法。基本思想是通过电路的特性分析和灵敏度的计算,进行故障样本的分类及优化,再根据灵敏度的计算结果提取相应特征信息。以此构造故障样本特征集,然后作为BP神经网络的输入对网络进行训练与诊断。对滤波器的仿真结果表明,该方法构造的样本集训练出来的神经网络,对模拟电路故障诊断的平均正确率为85%,优于传统方法。     

基于广域录波数据和FCM聚类的电网故障诊断方法

断路器和保护设备存在误动和拒动的可能性,不正确的动作会引入错误的故障信息干扰电网故障诊断。针对上述问题提出了一种基于广域故障录波数据和模糊C均值(FCM)聚类的电网故障诊断方法。对母线、输电线路和变压器这3种电网元件的故障录波数据进行挖掘分析,从中提取能够有效区分故障元件的特征;建立基于模糊C均值聚类的电网故障元件诊断模型。对故障后的元件进行聚类分析,并计算各元件的故障可信度,最终确定故障元件。新英格兰39节点系统的仿真和实际算例表明所提方法能够快速准确定位故障元件且不受故障类型和位置的影响。