Variable reconstruction and sensor fault identification using canonical variate analysis

The detection and identification of faults in dynamic continuous processes has received considerable recent attention from researchers in academia and industry. In this paper, a canonical variate analysis (CVA)-based sensor fault detection and identification method via variable reconstruction is described. Several previous studies have shown that CVA-based monitoring techniques can effectively detect faults in dynamic processes. Here we define two monitoring indices in the state and noise spaces for fault detection and, for sensor fault identification, we propose three variable reconstruction algorithms based on the proposed monitoring indices. The variable reconstruction algorithms are based on the concepts of conditional mean replacement and object function minimization. The proposed approach is applied to a simulated continuous stirred tank reactor and the results are compared to those obtained using the traditional dynamic monitoring technique, dynamic principal component analysis (PCA). The results indicate that the proposed methodology is quite effective for monitoring dynamic processes in terms of sensor fault detection and identification.     

无线传感器网络中节点故障诊断方法的研究

无线传感器网络中故障节点会产生并传输错误数据,这将消耗节点的能量和带宽,同时会形成错误的决策。利用节点感知数据的空间相似性,提出了节点故障诊断的算法,通过对邻节点所感知的传感数据进行比较,从而确定检测节点的状态,并将测试状态向网络中其他相邻节点进行扩散。该算法对实现故障节点的检测具有较好的性能,实验结果验证了算法的可行性和有效性。     

双余度传感器缓变故障检测的改进残差检验法

分析了基于卡尔曼滤波器的残差检验法对传感器缓变故障检测的不敏感性原因.针对双余度传感器缓变故障检测,采取了先故障判断后故障定位的故障检测策略,并提出了一种基于移动伪正常状态的残差构造方法.数学仿真验证了改进方法比传统方法更能够及时准确地识别双余度传感器缓变故障.     

MRNN:一种新的基于改进型递归神经网络的WSN动态建模方法:应用于故障检测

提出了一种适用于无线传感器网络WSN的故障检测方法,该方法运用改进的递归神经网络MRNN为WSN的节点、节点的动态特性以及节点间的关系建立相关模型,对WSN节点进行识别和故障检测。MRNN的输入选择建模节点的先前输出值及其邻居节点的当前及先前输出值,模型基于一种新的改进的反向传播型神经网络,该神经网络的输入以及传感器网络的拓扑结构基于通用的非线性传感器模型。仿真实验将MRNN方法与卡尔曼滤波法进行了全面的比较。实验表明,MRNN在置信因子较小的情况下与卡尔曼滤波方法相比有较高的故障检测精度。     

基于稀疏分解残差的氢气传感器故障探测与辨识方法

针对传感器故障探测和诊断,提出了一种基于稀疏分解残差的氢气传感器故障探测和辨识方法.基于信号稀疏分解理论,对采集的传感器正常信号数据集,利用K奇异值分解(K-SVD)学习算法得到一超完备字典D;在字典上对非正常(故障)信号进行分解,根据稀疏分解的残差大小和范围完成对传感器故障的探测及辨识.实验结果表明:对氢气传感器的故障探测率和总辨识率分别达到98.75%和97.25%,可以有效地解决氢气传感器的故障探测和辨识.     

Set-membership identification and fault detection using a Bayesian framework

This paper deals with the problem of set-membership identification and fault detection using a Bayesian framework. The paper presents how the set-membership model estimation problem can be reformulated from the Bayesian viewpoint in order to, first, determine the feasible parameter set in the identification stage and, second, check the consistency between the measurement data and the model in the fault-detection stage. The paper shows that, assuming uniform distributed measurement noise and uniform model prior probability distributions, the Bayesian approach leads to the same feasible parameter set than the well-known set-membership technique based on approximating the feasible parameter set using sets. Additionally, it can deal with models that are nonlinear in the parameters. The single-output and multiple-output cases are addressed as well. The procedure and results are illustrated by means of the application to a quadruple-tank process.     

Structured residual vector-based approach to sensor fault detection and isolation

This paper proposes a novel approach to detection and isolation of faulty sensors in multivariate dynamic systems. After formulating the problem of sensor fault detection and isolation in a dynamic system represented by a state space model, we develop the optimal design of a primary residual vector for fault detection and a set of structured residual vectors for fault isolation using an extended observability matrix and a lower triangular block Toeplitz matrix of the system. This work is, therefore, a vector extension to the earlier scalar-based approach to fault detection and isolation. Besides proposing a new algorithm for consistent identification of the Toeplitz matrix from noisy input and output observations without identifying the state space matrices {A, B, C, D} of the system, the main contributions of this newly proposed fault detection and isolation scheme are: (1) a set of structured residual vectors is employed for fault isolation; (2) after determination of the maximum number of multiple sensors that are most likely to fail simultaneously, a unified scheme for isolation of single and multiple faulty sensors is proposed; and (3) the optimality of the primary residual vector and the structured residual vectors is proven. We prove the advantage of our newly proposed vector-based scheme over the existing scalar element-based approach for fault isolation and illustrate its practicality by simulated and experimental evaluation on a multivariate pilot scale, computer interfaced system.     

Survey and application of sensor fault detection and isolation schemes

Model-based sensor fault detection, isolation and accommodation (SFDIA) is a direction of development in particular with UAVs where sensor redundancy may not be an option due to weight, cost and space implications. SFDIA via neural networks (NNs) have been proposed over the years due to their nonlinear structures and online learning capabilities. The majority of papers tend to consider single sensor faults. While useful, this assumption can limit application to real systems where sensor faults can occur simultaneously or consecutively. In this paper we consider the latter scenario, where it is assumed that a 1 s time gap is present between consecutive faults. Furthermore few applications have considered fixed-wing UAVs where full autonomy is most needed. In this paper an EMRAN RBF NN is chosen for modelling purposes due to its ability to adapt well to nonlinear environments while maintaining high computational speeds. A nonlinear UAV model is used for demonstration, where decoupled longitudinal motion is considered. System and measurement noise is also included in the UAV model as wind gust disturbances on the angle of attack and sensor noise, respectively. The UAV is assumed to operate at an initial trimmed condition of speed, 32 m/s and altitude, 1000 m. After 30 separate SFDIA tests implemented on a 1.6 GHz Pentium processor, the NN-SFDIA scheme detected all but 2 faults and the NN processing time was 97% lower than the flight data sampling time.     

基于加权变换的传感器故障检测新方法

考虑到传感器不可能具有等准确度这一客观事实,提出了一种改进的均值-偏差积检测法,即通过加权变换的方法将不等准确度的传感器,根据其准确度分别赋予不同的权值,然后组成综合表决系统,进行多传感器冗余系统的故障检测。该方法不受传感器等准确度和故障传感器输出不能相同等假设条件的限制,既减弱了使用的假设条件,扩大了应用范围;又提高了检测的正确率,有利于提高数据融合准确度和降低设备成本。     

Robust detection of intermittent multiplicative sensor fault

In this paper, the detection problem of intermittent multiplicative sensor fault is investigated for stochastic uncertain systems. A robust optimal filter is designed according to the criterion of minimum estimation error covariance. Then, based on this, a residual generator is constructed, and the quantitative effect of the fault on it is discussed in detail. After that we design the evaluation function and detection threshold to achieve intermittent fault detection. Our proposed strategy has a recursive form and only includes simple arithmetic operations, thus it is suitable for real‐time online applications. Finally, a simulation example is given to demonstrate the effectiveness of the proposed strategy.     

不等长批次过程的有序时段划分、建模及故障检测

对具有不等长时段的多时段批次过程进行监测是十分重要而且具有一定难度的. 时段在批次间的错位现象导致时间方向的不同过程特性混合在一起, 这给时段分析以及在线应用带来了一系列的问题. 为了解决不等长所带来的问题, 本文提出一种基于不等长时段有序识别及建模的故障检测方法. 该方法的主要贡献包括以下方面：1) 该方法通过步进地衡量过程的变量相关性对模型精度以及监测性能的影响, 自动有序地识别出每个不等长时段; 2) 在每个时段内,通过对不规则的过程数据进行整合建立了时段模型以捕捉不规则的时段特性; 3) 本文提供了一种简单而有效的在线判断新样本隶属时段和监测其运行状态的方法. 最后, 本文通过一个实例–具有不等长批次长度的注塑过程阐述了本方法的有效性.     

基于故障重构的故障识别

针对基于主元分析(Principal Component Analysis,PCA)的统计过程性能监测法,尽管不依赖于精确的数学模型,然而却限制了其在故障诊断方面的能力问题,在故障重构技术的基础上,研究了基于统计量的故障诊断问题,获得了主元子空间中故障可重构性的理论条件,提出了故障识别指标和诊断算法.通过对双效蒸发过程...     

Distributed fault detection over sensor networks with Markovian switching topologies

This paper deals with the distributed fault detection for discrete-time Markov jump linear systems over sensor networks with Markovian switching topologies. The sensors are scatteredly deployed in the sensor field and the fault detectors are physically distributed via a communication network. The system dynamics changes and sensing topology variations are modeled by a discrete-time Markov chain with incomplete mode transition probabilities. Each of these sensor nodes firstly collects measurement outputs from its all underlying neighboring nodes, processes these data in accordance with the Markovian switching topologies, and then transmits the processed data to the remote fault detector node. Network-induced delays and accumulated data packet dropouts are incorporated in the data transmission between the sensor nodes and the distributed fault detector nodes through the communication network. To generate localized residual signals, mode-independent distributed fault detection filters are proposed. By means of the stochastic Lyapunov functional approach, the residual system performance analysis is carried out such that the overall residual system is stochastically stable and the error between each residual signal and the fault signal is made as small as possible. Furthermore, a sufficient condition on the existence of the mode-independent distributed fault detection filters is derived in the simultaneous presence of incomplete mode transition probabilities, Markovian switching topologies, network-induced delays, and accumulated data packed dropouts. Finally, a stirred-tank reactor system is given to show the effectiveness of the developed theoretical results.     

Subspace identification of continuous time models for process fault detection and isolation

This paper proposes a novel subspace approach towards identification of optimal residual models for process fault detection and isolation (PFDI) in a multivariate continuous-time system. We formulate the problem in terms of the state space model of the continuous-time system. The motivation for such a formulation is that the fault gain matrix, which links the process faults to the state variables of the system under consideration, is always available no matter how the faults vary with time. However, in the discrete-time state space model, the fault gain matrix is only available when the faults follow some known function of time within each sampling interval. To isolate faults, the fault gain matrix is essential. We develop subspace algorithms in the continuous-time domain to directly identify the residual models from sampled noisy data without separate identification of the system matrices. Furthermore, the proposed approach can also be extended towards the identification of the system matrices if they are needed. The newly proposed approach is applied to a simulated four-tank system, where a small leak from any tank is successfully detected and isolated. To make a comparison, we also apply the discrete time residual models to the tank system for detection and isolation of leaks. It is demonstrated that the continuous-time PFDI approach is practical and has better performance than the discrete-time PFDI approach.     

分布式无线传感器网络故障检测算法综述

阐述了传感器网络中节点发生故障的原因并建立了故障模型。将当前主要的分布式节点故障检测算法分成了基于多数投票策略、基于中值策略、基于决策扩散策略、基于加权和基于分簇的算法五大类,详细阐述了分布式无线传感器网络故障检测算法的原理和步骤,并指出了各个算法的优势与不足。最后,对各个算法的性能进行了分析与比较,讨论了算法存在的问题,并指出了进一步的研究方向。     

Intermittent sensor fault detection for stochastic LTV systems with parameter uncertainty and limited resolution

ABSTRACT

This paper considers the detection problem of intermittent sensor faults in stochastic linear time-varying systems with both parameter uncertainty and limited resolution. By introducing the soft measurement model, a state estimator is designed whose upper bound of estimation error covariance is obtained and minimised at each time step. Based on it, the residual is generated and its relationship with the fault is analysed quantitatively. Then the evaluation function and corresponding detection threshold is given. Our proposed method is recursive and therefore suitable for real-time online applications. At last, two simulation studies are carried out to illustrate the validity of our proposed method.     

SR-30 turbojet engine real-time sensor health monitoring using neural networks,and Bayesian belief networks

This paper describes the use of artificial intelligence-based techniques for detecting and isolating sensor failures in a turbojet engine. Specifically, three artificial intelligence (AI) techniques are employed: artificial neural networks (NNs), statistical expectations, and Bayesian belief networks (BBNs). These techniques are combined into an overall system that is capable of distinguishing between sensor failure and engine failure—a critical capability in the operation of turbojet engines. The turbojet engine used in this study is an SR-30 developed by Turbine Technologies. Initially, NNs were designed and trained to recognize sensor failure in the engine. The increased random noise output from failing sensors was used as the key indicator. Next, a Bayesian statistical method was used to recognize sensor failure based on the bias error occurring in the sensors. Finally, a BBN was developed to interpret the results of the NN and statistical evaluations. The BBN determines whether single or multiple sensor failures signify engine failure, or whether sensor failures represent separate, unrelated incidences. The BBN algorithm is also used to distinguish between bias and noise errors on sensors used to monitor turbojet performance. The overall system is demonstrated to work equally well during start-up and main-stage operation of the engine. Results show that the method can efficiently detect and isolate single or multiple sensor failures within this dynamic environment.     

Scaled conjugate gradient and Bayesian training of neural networks for fault identification in cylinders

This paper compares Bayesian training of neural networks using hybrid Monte Carlo to scaled conjugate gradient method for fault identification in cylinders using vibration data. From the measured data pseudo-modal energies and modal properties are calculated and the coordinate pseudo-modal energy assurance criterion (COMEAC) and the coordinate modal assurance criterion (COMAC) are computed respectively. The pseudo-modal energies, modal properties, COMEAC and COMAC are used to train four neural networks. On average, the pseudo-modal-energy-networks are more accurate than the modal-property-networks. The weighted averages of the pseudo-modal-energy- and modal-property-networks form a committee of networks. The committee method gives lower mean squared errors and better classification of faults than the individual methods. The Bayesian training is found to be more accurate and computationally expensive than the scaled conjugate gradient method and to give confidence levels.     

Structural analysis for the sensor location problem in fault detection and isolation

In this paper we tackle the sensor location problem for fault detection and isolation based on structural analysis for linear systems with faults. We deal with this problem when the system under consideration is structured, that is, the entries of the system matrices are either fixed zeros or free parameters. With such structured systems one can associate a graph. A dedicated residual set is designed using a bank of observers for solving the problem. A major tool in this analysis is the notion of input separator in the associated graph, these separators form a lattice structure. The main contribution of this paper is the formulation of necessary and sufficient solvability conditions for the problem in terms of number of additional sensors measuring variables between faults and input separators in the associated graph.