Parallel processing application in traction motor fault diagnosis

Parallel processing-based system condition monitoring and fault detection are the order of the day for providing greater safety and reliability of the system. Fast and accurate fault diagnosis and detection predictor provides better system protection and maintenance planning of spare parts. In this paper, an attempt has been made to predict the temperature profile of different parts of a traction motor by using parallel processing techniques. The prediction of temperature of different parts of the motor during healthy and faulty condition have been established by developing a thermal model of the machine and transputer based concurrent process predictor has been taken up to assess the thermal behaviour of the machine The significant advantage of this predictor is its capability of providing early warning about the condition of the machine with greater confidence and built in redundancy and thereby ensured the protection of the machine.     

Design of testability for analogue fault diagnosis

With the rapidly increasing complexity of circuits and systems, the ability adequately to design a diagnosable circuit or system is a prime requisite for rapid fault location. In the present paper, a necessary and sufficient condition for pseudo-circuit generation is presented for the use of the pseudo-circuit in the self-test algorithm. to determine the diagnosability of a designed circuit, a testability condition is presented with examples. the unique feature of these conditions is that they depend only on the topological structure of a given circuit, not on the component values; therefore, these conditions can be implemented in both linear and non-linear circuits or systems. With the aid of a computer, the test points may be generated automatically by the proposed algorithm. Based on the proposed algorithm, the design of testability can be established.     

The relaxation pseudocircuit method for analogue fault diagnosis

An efficient frequency domain relaxation frequency pseudocircuit (RPC) approach is proposed and the associated solvability conditions are derived for analogue fault diagnosis. In this approach all the elements in the circuit are described by the associated relaxation models and the solution of the RPC is obtained iteratively by the phaseor technique. Consequently the diagnostic process is unified in the frequency domain and both the development of the theory and the implementation for practical application are made easier via this unified approach. Besides, both the order of the RPC and the measurement cost are reduced. This is useful especially when a high-Q network is involved or the solution is hard to obtain in the time domain. Two examples are illustrated to show the efficacy of the proposed method.     

An algorithm for multiple fault diagnosis in analogue circuits

The paper deals with multiple fault diagnosis of analogue AC or DC circuits with limited accessible terminals for excitation and measurement and brings an algorithm for identificating faulty elements and evaluating their parameters. The main achievement is a method enabling us to efficiently identify faulty elements. For this purpose some testing equations are derived playing a key role in identification of possibly faulty elements which are next verified using a test of acceptance. The proposed approach is described in detail for double fault diagnosis. Also extension to triple fault diagnosis is given. Although the method pertains to linear circuits, some aspects of multiple fault diagnosis of non‐linear circuits can be also performed using the small signal approach. Two numerical examples illustrate the proposed method and show its efficiency. Copyright © 2006 John Wiley & Sons, Ltd.     

A new approach for automatic fault diagnosis in analogue circuits

In this paper the problem of fault diagnosis in analogue circuits is addressed by a method which can be regarded as an artificial intelligence approach. the topology of the circuit under test along with the nominal element values, the necessary element models and their possible failure modes are the basic tools used in the procedure. Fault location is accomplished using appropriate DC node voltage measurements under certain DC excitations on the nominal circuit and the faulty one. the detection of signal faults which have DC effects is the goal of the method; cases of faults with non-DC effects are also examined. Demonstrative examples are given to show the applicability of the method on passive and active analogue circuits.     

基于MapReduce的电力设备并行故障诊断方法

以智能电网中电力变压器故障诊断为例,给出了基于MapReduce的电力变压器并行故障诊断过程,其应用4个MapReduce过程执行故障诊断算法的训练阶段,并得出分类模型,应用1个MapReduce过程完成对电力设备状态信息数据的故障诊断。建立了电力设备状态信息并行故障诊断实验平台,基于海量变压器油中溶解气体分析数据进行并行故障诊断实验,实验结果表明并行故障诊断速度高于传统单机环境下的诊断速度,满足智能电网环境下对海量电力设备状态信息快速故障诊断的要求。     

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.     

Parallel processing

Parallel processing is many problem solvers operating simultaneously or concurrently to reach a solution. A parallel processing computer system typically breaks down a big problem into many subproblems. It then solves it through close cooperation of a large number of interconnected processors. Parallel processing is the only way to overcome the limitation of traditional architectures. Following this direction, many new computer architectures have been proposed. They have opened up the possibility of constructing massively parallel computer systems that provide ultra high performance at reasonable costs     

基于多类AdaBoost的故障诊断算法

主要研究了多类分类AdaBoost算法,及其在多类故障诊断问题中的应用.为了解决“一对一”算法和“一对余”算法的局限性,提出了基于决策树的AdaBoost算法.利用遗传算法的全局随机搜索性能对数据集进行特征筛选,得到新的特征数据集,根据CART算法构造决策树建立AdaBoost分类器,使得决策树每一个节点的最可分类别尽...     

Research on fault diagnosis for MMC-HVDCSystems

Introduction: With the development of flexible HVDC technology, the fault diagnosis of MMC-HVDC becomes a new research direction. Based on the fault diagnosis theory, this paper proposes a robust fault diagnosis method to study the fault diagnosis problem of MMC-HVDC systems. Methods: By optimizing the gain matrix in the fault observer, fault detection with good sensitivity and robustness to disturbance is achieved. In the MMC-HVDC system, because of the inherently uncertain system and the presence of various random disturbances, the study of robust fault diagnosis method is particularly important. Results: Simulation studies during various AC faults have been carried out based on a 61-level MMC-HVDC mathematical model. The results validate the feasibility and effectiveness of the proposed fault diagnosis method. Conclusions: So this fault diagnosis method can be further applied to the actual project, to quickly achieve system fault diagnosis and accurately complete fault identification.     

基于改进SMOTE不均衡样本处理和IHPO-DBN的变压器故障诊断方法研究

针对由于变压器故障样本不均衡和故障模型陷入局部最优而导致的分类准确率低的问题,提出了基于改进的合成少数类过采样技术和优化深度置信网络(deep belief network, DBN)的变压器故障诊断方法。首先采用聚类融合的K-means算法,通过分簇和匹配的方式筛选出不稳定的少数类样本用以改进中心点合成少数类过采样技术(center point synthetic minority oversampling technique, CP-SMOTE)算法,并对少数类样本进行扩增,解决了变压器故障数据分布不均衡的问题。其次,通过加入随机逆向学习和自适应惯性权重技术对猎食者优化算法进行改进,并用改进后的算法对DBN的内部参数进行优化调整,提高了模型精度。最后,将不同数据预处理情况下以及不同数据规模下的变压器故障模型进行仿真对比。结果表明,经过数据预处理和模型优化后的变压器故障识别准确率能够提高到98%,有效地解决了故障数据不平衡导致的分类精度低的问题。     

飞行器自动故障诊断工具

目前大部分C-5飞行器存在大量的维修问题,据统计,每22个C-5飞行器就有8个不能正常执行任务,因此改进C-5飞行器的故障诊断系统非常必要.文中介绍改进后的MADARSⅢ战场维修故障诊断系统将极大地增强故障定位精度.MADARS在飞行器工作期间使用计算机记录上百个测试点的数据,测试点自动将故障代码对识别的故障元件自动进行触发,许多元件故障需要进一步隔离以便维修,MADARS通过通用处理接口系统(GPIS)下载数据,并进行有效数据的分析.GPIS还将数据传输到数据库服务器(QLD),测试点数据可以通过QLD存取,绘制曲线,与相关测试点比较,并出具报告,供维修人员使用.     

盲信号处理在感应电动机复合故障诊断中的应用研究

感应电动机定子短路和轴承故障一起发生时,由于故障机理不同,通常需要同时采集定子电流和振动信号才能进行有效诊断。研究了一种通过振动信号同时诊断两种故障的方法,以两种故障一起发生时的低频振动信号为研究对象,采用最小均方盲提取算法对故障信号进行提取,得到了信噪比提高的故障信号,做频谱分析后两种故障的特征频谱能够很好地分辨。通过实验平台模拟了定子相间短路和轴承外滚道缺损的复合故障,诊断结果表明该方法能够提取故障信号,减少噪声干扰,实现了利用振动信号对感应电动机复合故障的诊断。     

Improved analogue fault coverage estimation using probabilistic analysis

A method for fault detection probability estimation using statistical multi‐parameter circuit simulation is proposed, in order to check circuits for which double or multiple analogue measurements are utilized. Theoretical analysis for the estimation of the fault coverage is given, based on conditional probability calculations. The proposed method can be applied for both test measurement and input stimulus selection. Simulation results from the application of the method on typical analogue circuits—filter and amplifier—are given, showing a sufficient improvement over the fault coverage achieved by single measurements. Copyright © 2009 John Wiley & Sons, Ltd.     

Basic vibration signal processing for bearing fault detection

Faculty in the College of Engineering at the University of Alabama developed a multidisciplinary course in applied spectral analysis that was first offered in 1996. The course is aimed at juniors majoring in electrical, mechanical, industrial, or aerospace engineering. No background in signal processing or Fourier analysis is assumed; the requisite fundamentals are covered early in the course and followed by a series of laboratories in which the fundamental concepts are applied. In this paper, a laboratory module on fault detection in rolling element bearings is presented. This module is one of two laboratory modules focusing on machine condition monitoring applications that were developed for this course. Background on the basic operational characteristics of rolling element bearings is presented, and formulas given for the calculation of the characteristic fault frequencies. The shortcomings of conventional vibration spectral analysis for the detection of bearing faults is examined in the context of a synthetic vibration signal that students generate in MATLAB. This signal shares several key features of vibration signatures measured on bearing housings. Envelope analysis and the connection between bearing fault signatures and amplitude modulation/demodulation is explained. Finally, a graphically driven software utility (a set of MATLAB m-files) is introduced. This software allows students to explore envelope analysis using measured data or the synthetic signal that they generated. The software utility and the material presented in this paper constitute an instructional module on bearing fault detection that can be used as a stand-alone tutorial or incorporated into a course.     

VSC-HVDC送端换流器IGBT开路故障诊断方法

为了实现VSC-HVDC系统送端换流器IGBT开路故障的在线诊断,首先基于送端换流器拓扑结构得到了换流器开关函数模型,给出了基于开关函数的三相电压残差定义;然后推导了2类单桥臂IGBT开路故障和4类双桥臂IGBT开路故障下的三相电压残差计算公式,分析了IGBT开路故障的电压残差特征.基于此,通过设定电压残差阈值和IGB...     

A fault diagnosis method for three-phase rectifiers

In this article, we proposed a novel method based on principal component analysis and support vector machines was presented for fault diagnosis of three-phase rectifiers, in which the principal component analysis of fault signal is used to extract the features corresponding to various fault, then fault types are identified through the pattern recognition classifier based on support vector machines. The simulation result of fault diagnosis of a thyristor in a three-phase full-bridge controlled rectifier shows that the method can make an accurate identification of fault types as well as the location of the fault elements for power electronics circuits, and it has an excellent performance for noise robustness and calculation complexity. Therefore, it is quite practically valuable in the solution to the fault problems for power electronics rectifiers.     

改进深度置信网络对TE过程故障诊断研究

为了实现对TE过程的故障诊断，改进了深度置信网络（DBN）的故障诊断方法。传统DBN在训练过程有冗余特性，减弱网络的特征提取能力，改进DBN在无监督学习阶段的似然函数中加入惩罚正则项，通过稀疏约束得到DBN训练集的稀疏分布，再用Laplace函数的分布引导DBN节点的稀疏状态，用Laplace函数中的位置参数控制稀疏的力度，使无标签的数据特征更加直观的表示出来，最后将改进DBN和传统DBN、BP神经网络的仿真实验结果进行对比。实验结果，证明改进的DBN在故障诊断方面优于传统DBN和BP神经网络，达到了最好的诊断准确度，具有很高的理论研究价值。     

Parallel processing in power systems

A brief status report is presented of the emerging computer architectures and their probable influence on power-system computing. Experiences with three groups of computers are reviewed, namely supercomputers, array processors and multiprocessor networks.