Ellipsoidal bounding set-membership identification approach for robust fault diagnosis with application to mobile robots

A robust fault diagnosis approach is developed by incorporating a set-membership identification (SMI) method. A class of systems with linear models in the form of fault related parameters is investigated, with model uncertainties and parameter variations taken into account explicitly and treated as bounded errors. An ellipsoid bounding set-membership identification algorithm is proposed to propagate bounded uncertainties rigorously and the guaranteed feasible set of faults parameters enveloping true parameter values is given. Faults arised from abrupt parameter variations can be detected and isolated on-line by consistency check between predicted and observed parameter sets obtained in the identification procedure. The proposed approach provides the improved robustness with its ability to distinguish real faults from model uncertainties, which comes with the inherent guaranteed robustness of the set-membership framework. Efforts are also made in this work to balance between conservativeness and computation complexity of the overall algorithm. Simulation results for the mobile robot with several slipping faults scenarios demonstrate the correctness of the proposed approach for faults detection and isolation (FDI).

A Novel Parsimonious Neurofuzzy Model Applied to Railway Carriage System Identification and Fault Diagnosis

In this paper, we suggest a novel parsimonious neurofuzzy model realized by RBFNs for railway carriage system identification and fault diagnosis. To overcome the curse of dimensionality resulting from high dimensional input variables, in our developed model the features extracted from the available observations are regarded as the input variables by adopting the higher-order statistics(HOS) technique. Such a constructed model is also applied to a practical railway carriage system, simulation results indicate that the developed neurofuzzy model possesses strong identification and fault diagnosis ability.

Method for Analog Circuit Soft-Fault Diagnosis and Parameter Identification Based on Indictor of Phase Deviation and Spectral Radius

The soft fault induced by parameter variation is one of the most challenging problems in the domain of fault diagnosis for analog circuits.A new fault location and parameter prediction approach for soft-faults diagnosis in analog circuits is presented in this paper.The proposed method extracts the original signals from the output terminals of the circuits under test（CUT） by a data acquisition board.Firstly,the phase deviation value between fault-free and faulty conditions is obtained by fitting the sampling sequence with a sine curve.Secondly,the sampling sequence is organized into a square matrix and the spectral radius of this matrix is obtained.Thirdly,the smallest error of the spectral radius and the corresponding component value are obtained through comparing the spectral radius and phase deviation value with the trend curves of them,respectively,which are calculated from the simulation data.Finally,the fault location is completed by using the smallest error,and the corresponding component value is the parameter identification result.Both simulated and experimental results show the effectiveness of the proposed approach.It is particularly suitable for the fault location and parameter identification for analog integrated circuits.

Fault Diagnosis Based on Fuzzy Support Vector Machine with Parameter Tuning and Feature Selection

This study describes a classification methodology based on support vector machines(SVMs),which offer superior classification performance for fault diagnosis in chemical process engineering.The method incorporates an efficient parameter tuning procedure(based on minimization of radius/margin bound for SVM's leave-one-out errors)into a multi-class classification strategy using a fuzzy decision factor,which is named fuzzy support vector machine(FSVM).The datasets generated from the Tennessee Eastman process(TEP)simulator were used to evaluate the clas-sification performance.To decrease the negative influence of the auto-correlated and irrelevant variables,a key vari-able identification procedure using recursive feature elimination,based on the SVM is implemented,with time lags incorporated,before every classifier is trained,and the number of relatively important variables to every classifier is basically determined by 10-fold cross-validation.Performance comparisons are implemented among several kinds of multi-class decision machines,by which the effectiveness of the proposed approach is proved.

Wear Fault Diagnosis of Machinery Based on Neural Networks and Gray Relationships

In this paper, the regular characteristic of -wear particles related to fault type of machines based on condition monitoring of reciprocal machinery is discussed. The typical -wear particles spectrum is established according to the equipment structure , friction and wear rule and the characteristic of 'wear particles; The identification technology of wear particles is proposed based on neural networks and a gray relationship ; an intelligent wear particles identification system is designed. The diagnosis example shows that this system can promote the accuracy and the speed of wear particles identification.

Novel Fault Diagnosis Scheme for HVDC System via ESO

A novel fault detection and identification(FDI)scheme for HVDC(High Voltage Direct Current Transmission)system was presented.It was based on the unique active disturbance rejection concept,where the HVDC system faults were estimated using an extended states observer(ESO).Firstly,the mathematical model of HVDC system was constructed,where the system states and disturbance were treated as an extended state.An augment HVDC system was established by using the extended state in rectify side and converter side,respectively.Then,a fault diagnosis filter was established to diagnose the HVDC system faults via the ESO theory.The evolution of the extended state in the augment HVDC system can reflect the actual system faults and disturbances,which can be used for the fault diagnosis purpose.A novel feature of this approach is that it can simultaneously detect and identify the shape and magnitude of the HVDC faults and disturbance.Finally,different kinds of HVDC faults were simulated to illustrate the feasibility and effectiveness of the proposed ESO based FDI approach.Compared with the neural network based or support vector machine based FDI approach,the ESO based FDI scheme can reduce the fault detection time dramatically and track the actual system fault accurately.What's more important,it needs not do complex online calculations and the training of neural network so that it can be applied into practice.

IGIgram:An Improved Gini Index-Based Envelope Analysis for Rolling Bearing Fault Diagnosis

The transient impulse features caused by rolling bearing faults are often present in the resonance frequency band which is closely related to the dynamic characteristics of the machine structure.Informative frequency band identification is a crucial prerequisite for envelope analysis and thereby accurate fault diagnosis of rolling bearings.In this paper,based on the ratio of quasi-arithmetic means and Gini index,improved Gini indices(IGIs)are proposed to quantify the transient impulse features of a signal,and their effectiveness and advantages in sparse quantification are confirmed by simulation analysis and comparisons with traditional sparsity measures.Furthermore,an IGI-based envelope analysis method named IGIgram is developed for fault diagnosis of rolling bearings.In the new method,an IGI-based indicator is constructed to evaluate the impulsiveness and cyclostationarity of the narrow-band filtered signal simultaneously,and then a frequency band with abundant fault information is adaptively determined for extracting bearing fault features.The performance of the IGIgram method is verified on the simulation signal and railway bearing experimental signals and compared with typical sparsity measures-based envelope analysis methods and log-cycligram.The results demonstrate that the proposed IGIs are efficient in quantifying bearing fault-induced transient features and the IGIgram method with appropriate power exponent can effectively achieve the diagnostics of different axle-box bearing faults.

Welch功率谱与卷积神经网络结合的滚动轴承故障诊断

针对滚动轴承故障诊断在小训练样本下和强噪声下无法取得高精度识别的问题,提出一种基于Welch功率谱结合卷积神经网络进行诊断的方法。该方法以原始时域振动信号作为输入,用Welch功率谱转换数据形态同时对高强度噪声进行抑制,再用得到的功率谱训练卷积神经网络,最后将训练好的模型用于轴承的故障诊断。与WDCNN[1]等方法进行对比,实验发现在混合负载下,该方法平均识别率正确达到99%,其它方法达到这个精度至少需要20倍以上的训练样本量,明显优于WDCNN等方法。抗噪实验结果表明噪声对信号的干扰越强烈,该方法的抗噪表现越好,其抗噪性能要显著优于WDCNN等方法。

Fault diagnosis of wind turbine bearing based on stochastic subspace identification and multi-kernel support vector machine

In order to accurately identify a bearing fault on a wind turbine, a novel fault diagnosis method based on stochastic subspace identification(SSI) and multi-kernel support vector machine(MSVM) is proposed. Firstly, the collected vibration signal of the wind turbine bearing is processed by the SSI method to extract fault feature vectors. Then, the MSVM is constructed based on Gauss kernel support vector machine(SVM) and polynomial kernel SVM. Finally, fault feature vectors which indicate the condition of the wind turbine bearing are inputted to the MSVM for fault pattern recognition. The results indicate that the SSI-MSVM method is effective in fault diagnosis for a wind turbine bearing and can successfully identify fault types of bearing and achieve higher diagnostic accuracy than that of K-means clustering, fuzzy means clustering and traditional SVM.

基于QAR数据的飞行控制系统故障研究综述

为系统梳理国内外对民用飞机飞行控制系统故障分析的研究历程和现状,针对基于快速存取记录器(QAR)数据分析的飞行控制系统典型故障类型,首先,总结QAR数据预处理、特征提取等使用过程;然后,根据故障分析可达到的性能指标,提出4个故障研究阶段,分别为故障监测、故障识别、故障诊断和故障预测;最后,综合国内外研究进度与深度,得出飞行控制系统典型故障类型,包括方向舵液压泄漏、升降舵指示不一致、襟翼动作耗时等,建模常用QAR数据项包括飞机主舵面位置、飞行姿态、飞机性能、左右襟翼角度、襟翼位置等,计算方法包括物理模型、多变量统计、逻辑推理、机器学习等。结果表明:系统分析方向舵、升降舵、襟翼等子系统最新研究进展,发现在故障类型、参数选择和计算方法的改进等方面取得了一定的成果,故障研究阶段基本处于故障诊断或非实时预测水平,但仍需加强面向安全保障与实际维修方面的需求,以实现故障实时预测技术。

基于优化矩阵扰动分析的模拟电路故障诊断

在现有的模拟电路故障诊断算法中,人工智能故障诊断算法训练数据量大、训练时间长,且难以实现参数辨识。传统电路分析方法所需测试点多,计算复杂。基于此,提出了一种基于优化矩阵扰动分析的模拟电路故障诊断算法。首先,采用拉普拉斯(Laplace)算子卷积被测电路的输出响应矩阵,从而增强矩阵元素与电路元件参数之间的扰动规律。其次,选取矩阵的迹和谱半径作为故障特征,并利用这种扰动规律建立矩阵模型。然后,利用改进的诊断算法,在Sallen_Key带通滤波器电路和跳蛙低通滤波器电路上进行实例验证。结果表明,所提方法在仅使用一个测点的情况下,可实现故障元件的参数辨识。其故障诊断率达100%,参数辨识误差控制在1%内,且计算时间控制在毫秒级别。因此该方法容易实现在线测试,且适用于要求高定位准确率、高精度参数辨识的场合。

基于随机子空间法的滑动轴承运行模态参数识别

滑动轴承的运行模态参数是其状态监测和早期故障诊断的重要指标,利用随机子空间法识别滑动轴承运行模态参数时,环境噪声和阶数过估计引起的虚假模态会影响真实模态参数的识别。为减少虚假模态的干扰,首先对振动信号利用互补集合经验模态分解和小波变换相结合的方法进行降噪处理,然后将预处理后的信号分段并分别进行模态参数识别,通过对比同阶极点获得更清晰的稳定图,最后采用谱聚类算法实现模态参数的自动选择。通过数值仿真和相关试验验证该方法的有效性。

故障自诊断技术在煤矿传感设备中的应用研究

矿用传感器是煤矿安全生产过程中最重要的设备之一,其智能化应用已经成为了煤矿行业传感类设备的发展趋势。由于各类矿用传感器的工作环境恶劣和长期运行,关键部件不可避免产生疲劳损坏,导致传感器的故障率相对较高;针对性介绍了故障自诊断技术在煤矿用传感层设备智能化应用中的研究现状和发展趋势,提出一种具有故障自诊断功能的矿用传感器的设计方法。该方法采用实时采集供电电源、模拟信号、数字信号等特征数据,构建了故障诊断的优先级识别模型,设计了故障自诊断技术的矿用传感器,增强了传感器的工作的可靠性,提升了传感器的智能化应用水平。传感器具备智能自诊断功能,对于保障煤矿生产安全和提高生产效率具有重要意义。

基于图谱功率谱熵和最大均值差异的GIS机械状态辨识方法

针对常规方法对于气体绝缘金属封闭开关设备(Gas Insulated Switchgear,GIS)机械缺陷的特征识别稳定性差、识别率低的问题,在图谱理论的基础上,提出一种基于图谱功率谱熵和最大均值差异(Maximum Mean Discrepancy,MMD)的GIS机械状态辨识方法。首先将采集得到的GIS振动信号转化为图信号,并利用图傅里叶变换技术变换至图谱域进行分析处理;然后提取图谱功率谱熵作为表征GIS不同状态的特征参数;最后利用MMD距离判别函数实现GIS不同工况下的状态辨识。实验结果表明:在噪声干扰的情况下,所提方法能够有效提取GIS不同状态下的特征参数,并成功区分出屏蔽罩松动及内部异物缺陷,状态辨识精度高达93.89%,较常规方法有明显提高。

HVD分解和GA-BP神经网络结合的井架钢结构损伤识别

针对井架钢结构冲击载荷振动信号非线性、非平稳性对损伤识别的干扰问题,提出了一种基于希尔伯特振动分解(Hilbert Vibration Decomposition,HVD)与遗传算法优化的神经网络(Genetic BP Neural Networks,GA-BP)相结合的智能故障诊断方法。首先,利用HVD分解的方法处理冲击载荷作用下的加速度非平稳振动信号;其次,由斯皮尔曼相关系数选取HVD分解后的最优(Intrinsic Mode Function,IMF)分量,以最优IMF分量能量变化率构造特征向量;最后,通过特征向量建立数据集进行神经网络训练,完成信号的特征学习和故障分类。利用ZJ70型井架钢结构模型进行冲击载荷作用下的单处损伤和多处损伤的不同工况实验验证,结果表明:对于单处损伤位置识别率达到90%,多处损伤位置识别率高达96%,利用HVD分解与GA-BP神经网络相结合的方法具有较好的稳定性,能够准确判断出井架钢结构损伤位置,具有一定的实际应用价值。

发动机故障信息综合研究平台构建技术研究

针对目前多型号多批次的无人机发动机数据存储较为分散、缺少专门的分析手段辅助设计人员快速分析发动机状态的问题,借鉴国内外较为成熟的故障诊断算法和平台构建技术,设计了发动机故障信息综合研究平台。平台内集成了飞行数据状态识别、故障诊断、分析数据库建立等多项关键技术,结合不同的飞行工况分类结果,从长期和短期的角度分别采取基于递归结构辨识(Recursive Structural Identification,RESID)的起动状态故障检测、基于自回归滑动平均(Auto Regressive Moving Average,ARMA)模型的稳态故障检测、基于参数趋势分析的故障诊断和基于深度学习的智能故障识别等不同的诊断方法,并采用飞行报告的形式实现了从数据文件输入至飞行状态分析的全流程计算和结果展示,有效地实现了发动机使用维护的数据支持和技术保障,具有一定的工业应用价值。

变电站电力设备声纹振动在线监测技术研究

声纹振动监测技术作为一种新兴的在线监测手段,已在变电站电力设备的运行状态监测中得到广泛应用。在过去几年里,随着传感器技术、信号处理技术和数据分析算法的不断发展,声纹振动监测技术取得了显著的进展。本文深入研究了变电站电力设备的声纹振动监测技术,探讨了系统架构故障诊断与预测方法,并设计了声纹振动在线监测系统。研究结果表明,声纹振动监测技术在变电站电力设备故障识别和预测方面具有很高的准确性和可靠性。

基于蝴蝶优化算法的电力系统故障诊断应用研究

针对电力系统故障诊断的不确定性,以及保护开关和断路器动作信息中常出现的误动、拒动等问题,笔者提出一种基于蝴蝶优化算法的电力系统故障诊断方法,重点分析了该方法在解析大数据模型中对目标问题快速求解的特点,给出了数据引入时的预处理过程,充分利用蝴蝶优化算法在寻迹方面的推理能力,根据报警信息搜索电网数据库,提出故障假设及诊断结果。采用IEEE-39节点进行验证,并与人工神经网络方法进行比较。结果表明,该方法目标更加明确,且收敛速度更快,适用于网络规模大的电力系统故障诊断。

频域特征驱动的车辆旋转部件灰色故障诊断

根据不同故障引起的振动信号频域特征各异的特点,首先,运用调Q小波变换(tunable Q-factor wavelet transform,简称TQWT)的频率响应对车辆旋转部件振动信号或其包络信号进行分析,构建以信号在不同频段各子带能量占比为元素的特征向量;其次,针对灰色接近关联度在处理2组相交的序列时存在的两序列变化趋势不同、原始累差小而导致关联度过大的问题,提出了灰色绝对接近关联度模型;最后,在所构建的频域特征向量驱动下,计算其与标准模式的灰色绝对接近关联度,对车辆关键旋转部件故障状态进行识别。利用所提方法对列车轮对轴承和汽车变速器齿轮箱不同运行状态的振动信号进行分析,结果表明,所提方法能够准确识别车辆旋转部件的运行状态和故障类型,通过对比分析验证了该方法的优越性。

基于状态监测的核电厂换热器完整性健康分析

设备冷却水换热器在核电厂中扮演着至关重要的角色,负责散热并确保核电厂的安全运行。为满足核电厂换热器健康状态监测的需求,提出了一种基于指数移动平均的监测参数数据处理方法,通过计算并提取各个指标的数据基线,有效剔除波动数据,保留监测参数的动态特征。通过锁定异常参数,并结合异常工况分析库,最终实现对换热器健康状态的分析。最后,通过泄漏试验验证了基于MSET的异常工况健康分析的可行性。试验结果表明:该模型能够识别异常数据和异常工况,并具有较低的误报率。