Reconstruction-based Contribution Research Articles

An Industrial Internet Application for Real-Time Fault Diagnosis in Industrial Motors

Being able to detect, identify, and diagnose a fault is a key feature of industrial supervision systems, which enables advance asset management, in particular, predictive maintenance, which greatly increases efficiency and productivity. In this paper, an Industrial Internet app for real-time fault detection and diagnosis is implemented and tested in a pilot scale industrial motor. Real-time fault detection and identification is based on dynamic incremental principal component analysis (DIPCA) and reconstruction-based contribution (RBC). When the analysis indicates that one of the vibration measurements is responsible for the fault, a convolutional neural network (CNN) is used to identify the unbalance or bearing fault type. The application was evaluated in its three functionalities: fault detection, fault identification, and fault identification of vibration-related faults, yielding a fault detection rate over 99%, a false alarm rate below 5%, and an identification accuracy over 90%. Note to Practitioners —This paper focuses on designing and evaluating a real-time fault diagnosis application in an industrial setup. To this end, this paper also tackles the problem of developing a methodology for implementing advanced state-of-the-art fault detection techniques in real machinery, following industry standards and using a modern informatics architecture. The application here developed uses a statistical data-driven fault diagnosis technique, hence it requires a training stage using historical data to learn patterns and estimate parameters. A proof of concept in fault diagnosis for industrial motors is given; however, it should be noted that both the methodology and the deployed architecture are scalable and flexible enough to facilitate the implementation in other industrial environments. The implementation here presented was deployed using only open-source tools, which allows evaluating this tool without incurring in high expenses.

Efficient Nonlinear Fault Diagnosis Based on Kernel Sample Equivalent Replacement

Contribution plots and reconstruction-based contribution (RBC) are efficient linear diagnosis tools in multivariate statistical process monitoring. Unfortunately, they cannot be directly applied to nonlinear fault diagnosis with kernel-based methods due to kernel function covers up the information of the original process variables. Although existing kernel gradient-based approaches have solved this problem to a certain extent, they are still far from suitable for practical applications because they require extremely huge amounts of computation. Their calculations cannot be obtained in a tolerable time unless expensive hardware costs are involved. This paper will thoroughly address this issue by revealing a hidden but important equivalent relationship between the variance-covariance matrix of a centralized process variables matrix and the centralized kernel matrix. Based on this relationship, the nonlinear detection index can be transformed into an explicit quadratic form of variables sample, such that contribution plots and RBC can be directly applied to kernel-based fault diagnosis with a very limited amount of computation, just as their usages in the linear cases. Simulation results obtained from two industrial examples demonstrate the effectiveness of the new method.

A novel fault identification and root-causality analysis of incipient faults with applications to wastewater treatment processes

To detect incipient faults and identify the corresponding root-cause in the wastewater treatment processes, this paper proposed a novel fault diagnosis framework. In this framework, we firstly proposed a new method, called Moving average residual difference reconstruction contribution plot (Mard-RCP). Mard-RCP can solve the defect of traditional reconstruction-based contribution plot (RCP), which is not only required to obtain the sensor fault direction in advance, but also usually suffers from the fault smearing effect. Also, this method can effectively improve the diagnosis accuracy by manipulating the signal-to-noise ratio properly. Secondly, to address the traditional issue of contribution plot to identify root-variables regardless of causality, a novel fault candidate variables selection method, termed as VS-R, together with Granger causal (GC) analysis is introduced to locate the root-variables of the fault. The proposed fault diagnosis framework is simulated and validated on the BSM1 simulation platform proposed by International Water Association and in a full-scale wastewater treatment plant.

Decision Support System for Cyber Attack Diagnosis in Smart Water Networks

Abstract This paper presents the design of a Decision Support System (DSS) for cyber attack diagnosis in Smart Water Networks. Cyber physical systems (CPS) integrate physical assets with embedded systems. Smart water networks are CPS formed by the distribution system of pipes together with sensors, actuators, programmable logic controllers (PLCs) and supervisory control and data acquisition (SCADA) system. The proposed strategy combines three signal processing tools: Principal Component Analysis, the Adaptive Exponential Weighted Moving Average Chart and the Reconstruction-based Contribution method. The proposal is tested with the data published in the BATtle of the Attack Detection ALgorithms (BATADAL) competition (http://www.batadal.net). The results show that the proposed DSS allows the detection of all the attacks from the test data set 3 with a satisfactory performance by using only data from the normal behavior (data set 1). The proposed DSS can also successfully identify the area of the network and the physical component(s) under attack for different scenarios (first three attacks).

Fault diagnosis for a kind of nonlinear systems by using model-based contribution analysis

For the purpose of fault detection and isolation (FDI), reconstruction-based contribution (RBC) analysis is carried out in a model-based way. A bank of adaptive observers are designed for a set of potential faults. From these observers, fault estimates and fault signatures are directly available, thus contribution functions are conveniently constructed to accomplish the FDI work. This integrated design of contribution analysis and adaptive observer takes advantages of both data-driven and model-based approaches, and the diagnosis performance is improved. Furthermore, quantitative isolability analysis is also studied by similarity measurement of the obtained fault signatures. Simulation study with a nonlinear unmanned aerial vehicle (UAV) model shows the effectiveness of the proposed method.

An Unsupervised Approach to Leak Detection and Location in Water Distribution Networks

Abstract The water loss detection and location problem has received great attention in recent years. In particular, data-driven methods have shown very promising results mainly because they can deal with uncertain data and the variability of models better than model-based methods. The main contribution of this work is an unsupervised approach to leak detection and location in water distribution networks. This approach is based on a zone division of the network, and it only requires data from a normal operation scenario of the pipe network. The proposition combines a periodic transformation and a data vector extension together with principal component analysis of leak detection. A reconstruction-based contribution index is used for determining the leak zone location. The Hanoi distribution network is employed as the case study for illustrating the feasibility of the proposal. Single leaks are emulated with varying outflow magnitudes at all nodes that represent less than 2.5% of the total demand of the network and between 3% and 25% of the node’s demand. All leaks can be detected within the time interval of a day, and the average classification rate obtained is 85.28% by using only data from three pressure sensors.

Fault diagnosis for processes with feedback control loops by shifted output sampling approach

Data-driven fault diagnosis of closed loop processes has been a challenge in the process control community. The issue of the interaction between the process model and the controller model exists in models directly identified from closed loop data, because for all the measured process outputs, no matter whether they are normal or faulty, they are fed back into the controllers so that the reconstruction-based contribution (RBC) as the fault diagnosis method has a severe fault smearing effect. This article proposes a novel sampling scheme which can significantly eliminate the adverse effect of modeling issues in feedback control. The identifiability condition of model parameters is satisfied in the new sampling framework so that the RBC recovers its efficiency even though the process runs under feedback control. Two benchmarks, a continuous stirred-tank heater process and the Tennessee Eastman challenge problem, are used to test the efficiency of the proposed method.

Isolating incipient sensor fault based on recursive transformed component statistical analysis

This paper considers the isolation problem of incipient sensor fault. Based on recursive transformed component statistical analysis (RTCSA), two different isolation methods are proposed. The first method is called subspace reconstruction, where elements in specific subspaces are eliminated, and then reconstructed by minimizing the reconstructed detection index. The faulty variable is determined by the least scaled reconstructed detection index. The second method is called subblock detection, which has less online computational complexity. The subblocks of the measurement matrix are sequentially selected in each sliding window to calculate the subblock detection indices, and the faulty variable is determined by the largest subblock detection margin. Compared with the existing isolation methods such as reconstruction-based contribution (RBC) and its variant termed as average residual-difference reconstruction contribution plot (ARdR-CP), the superior isolation performances of the proposed methods are illustrated by a numerical example as well as a simulation on a continuous stirred tank reactor.

A New Strategy of Adaptive Observer Based Fault Isolation

The problem of fault isolation (FI) for a kind of nonlinear discrete time varying (NDTV) systems is investigated. A bank of adaptive observers are designed for the considered system with different potential faults, and the isolation work is transformed into finding out which observer matches the current system best. The idea of reconstruction-based contribution (RBC) analysis is introduced to construct indicator variables for all of the observers and the one with the largest indicator variable is declared as the isolation result. In this method, knowledge of the system model is fully used, while the idea of contribution analysis is typically data-driven. Finally, simulation study is carried out with a nonlinear unmanned aerial vehicle (UAV) model, and the results demonstrate that the proposed method provides us with a new promising strategy for FI.

Concurrent Monitoring and Diagnosis of Process and Quality Faults with Canonical Correlation Analysis

Abstract Partial least squares and canonical correlation analysis are latent variable models suitable for quality-relevant monitoring based on process and quality data. Recently, concurrent monitoring schemes are proposed to achieve simultaneous process and quality monitoring. This paper defines and analyzes quality-relevant monitoring based on these popular latent structure modeling methods, and the associated quality-relevant monitoring statistics are defined. Additionally, contribution plots and reconstruction-based contribution diagnosis methods are developed for concurrent fault diagnosis. Multi-dimensional quality-relevant faults can be diagnosed in the same reconstruction framework. Finally, a detailed case study on Tennessee Eastman process is shown to illustrate the diagnosis of process and quality faults and the prognosis of quality-relevant faults.

An efficient FDIR strategy on nonlinear processes of self-validating multifunctional sensors

PurposeThis paper aims to enhance the reliability of self-validating multifunctional sensors.Design/methodology/approachAn effective fault detection, isolation and data recovery (FDIR) strategy by using kernel principal component analysis (KPCA) coupled with gray bootstrap and fault reconstruction methods.FindingsThe proposed FDIR strategy is able to the address fault detection, isolation and data recovery problem of self-validating multifunctional sensors efficiently.Originality/valueA KPCA-based model which can overcome the limitation of existing linear-based models is used to achieve the fault detection task. By using gray bootstrap method, the position of all faulty sensitive units can be calculated even under the multiple faults situation. A reconstruction-based contribution method is adopted to evaluate the amplitudes of the fault signals, and the fault-free output of the faulty sensitive units can be used to replace the fault output.

Data validation of multifunctional sensors using independent and related variables

To enhance the reliability of multifunctional sensors, a novel data validation strategy is presented by handling independent and related variables separately. The maximal information coefficient (MIC), which can measure the strength of the correlation between two variables, is applied to divide all variables of multifunctional sensors into related and independent. For one thing, the k-nearest neighbor (kNN) rule is introduced to accomplish fault detection and isolation of independent variables, and the grey predictive model GM(1,1), which has the advantages of low computation burden and high accuracy, is adopted to achieve data recovery of faulty independent variables. For another, the kernel principal component analysis (KPCA), which can handle possible non-linearity of data, is employed to realize fault detection of related variables. An iterative reconstruction-based contribution (IRBC) method is developed to isolate all faulty related variables, and data recovery of them are implemented using a fuzzy similarity (FS)-based reconstruction method based on the spatial correlations among related variables. An experimental system for multifunctional sensors is built to evaluate the proposed strategy, and the performance comparisons with its counterparts are also conducted.

A Novel Decentralized Process Monitoring Scheme Using a Modified Multiblock PCA Algorithm

Decentralized process monitoring based on purely data-based methods has recently gained considerable attention in multivariate statistical process monitoring circle. Although the process variables can be divided into several blocks automatically according to their statistical preferences, most of the existing multiblock modeling strategies tends to build local monitoring models individually, where the relevance among different blocks is ignored, and this leaves a room for enhancing process monitoring performance. Inspired by the recognition of this lack, a modified multiblock principal component analysis (MBPCA) algorithm is proposed for extracting block scores with respect to both specificity in each block and relevance among different blocks. Based on this sort of modeling strategy, a novel decentralized process monitoring is formulated by incorporating a PCA-based process decomposition strategy for block division, Bayesian inference to achieve decision fusion of fault detection, and reconstruction-based contribution plots for fault diagnosis. The superiority and validity of the proposed method is finally demonstrated through comparison studies on two simulated examples.

Joint Data-Driven Fault Diagnosis Integrating Causality Graph With Statistical Process Monitoring for Complex Industrial Processes

In this paper, an integrated fault diagnosis method is proposed to deal with fault location and propagation path identification. A causality graph is first constructed for the system according to the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a priori</i> knowledge. Afterward, a correlation index (CI) based on the partial correlation coefficient is proposed to analyze the correlation of variables in causality graph quantitatively. To achieve accurate fault detection results, the proposed CI is monitored by probability principal component analysis. Moreover, the concept of weighted average value is introduced to identify fault propagation path based on reconstruction-based contribution and causality graph after detecting a fault. Finally, the new proposed scheme would be practiced with real industrial HSMP data, where the individual steps as well as the complete framework were extensively tested.

Normalized Relative RBC-Based Minimum Risk Bayesian Decision Approach for Fault Diagnosis of Industrial Process

In the industrial process, it is helpful to take the previous fault diagnosis results into consideration during the current determination of faulty variables. In this paper, an unsupervised data-driven fault diagnosis method based on normalized relative reconstruction-based contribution (RBC) and the minimum risk Bayesian (MRB) decision theory is presented. Normalized relative RBC is used to represent the characteristic of the observation of the samples, and beta distribution is adopted to approximate the probability density distribution of the variable being faulty or normal. On the basis of the adjustable loss function, the conditional risk is obtained. The fault diagnosis method based on MRB decision is proposed, which will reduce the influence of smearing effect, improve the diagnosis rate, handle faults with a small magnitude, and identify multiple process faults. Numerical simulation examples and the Tennessee Eastman process are given to show the effectiveness and superiority of the proposed method.

Statistical process monitoring based on orthogonal multi-manifold projections and a novel variable contribution analysis

Multivariate statistical methods have been widely applied to develop data-based process monitoring models. Recently, a multi-manifold projections (MMP) algorithm was proposed for modeling and monitoring chemical industrial processes, the MMP is an effective tool for preserving the global and local geometric structure of the original data space in the reduced feature subspace, but it does not provide orthogonal basis functions for data reconstruction. Recognition of this issue, an improved version of MMP algorithm named orthogonal MMP (OMMP) is formulated. Based on the OMMP model, a further processing step and a different monitoring index are proposed to model and monitor the variation in the residual subspace. Additionally, a novel variable contribution analysis is presented for fault diagnosis by integrating the nearest in-control neighbor calculation and reconstruction-based contribution analysis. The validity and superiority of the proposed fault detection and diagnosis strategy are then validated through case studies on the Tennessee Eastman benchmark process.

Data-driven Fault Detection and Diagnosis for HVAC water chillers

Faulty operations of Heating, Ventilation and Air Conditioning (HVAC) chiller systems can lead to discomfort for the users, energy wastage, system unreliability and shorter equipment life. Faults need to be early diagnosed to prevent further deterioration of the system behaviour and energy losses. Since it is not a common practice to collect historical data regarding unforeseen phenomena and abnormal behaviours for HVAC installations, in this paper, a semi-supervised data-driven approach is employed for fault detection and isolation that makes no use of a priori knowledge about abnormal phenomena. The proposed method exploits Principal Component Analysis (PCA) to distinguish anomalies from normal operation variability and a reconstruction-based contribution approach to isolate variables related to faults. The diagnosis task is then tackled by means of a decision table that associates the influence of faults to certain characteristic features. The Fault Detection and Diagnosis (FDD) algorithm performance is assessed by exploiting experimental datasets from two types of water chiller systems.

Incipient Sensor Fault Diagnosis Using Moving Window Reconstruction-Based Contribution

Reconstruction-based contribution (RBC) is widely used for fault isolation and estimation in conjunction with principal component analysis (PCA)-based fault detection. Correct isolation can be guaranteed by RBC for single-sensor faults with large magnitudes. However, the incipient sensor fault diagnosis problem is not well handled by traditional PCA and RBC methods. In this paper, the limitations of traditional PCA and RBC methods for incipient sensor fault diagnosis are illustrated and analyzed. Through the introduction of a moving window, a new strategy based on the PCA model is presented for incipient fault detection. Regarding incipient fault isolation and estimation, a new contribution analysis method called moving window RBC is proposed to enhance the isolation performance and estimation accuracy. Rigorous fault detectability and isolability analyses of the proposed methods are provided. In addition, effects of the window width on fault detection, isolation, and estimation are discussed. Simulation ...

On the use of reconstruction-based contribution for fault diagnosis

In the multivariate statistical process monitoring (MSPM) area, principal component analysis (PCA) and reconstruction-based contribution (RBC) are two commonly used techniques for fault detection and fault diagnosis problems, respectively. This paper starts with a review of the two methods. It is then pointed out that, when the dimensionality of the principal component subspace or the residual subspace in the PCA model is equal to 1, several fault detection indices based RBC will be invalid for fault diagnosis. Corresponding geometric interpretations of the invalidation cases are illustrated intuitively according to the definition of RBC. In order to perform effective fault diagnosis in such invalidation cases, three methods including the available combined index based RBC, the derived Mahalanobis distance based RBC, and the proposed chi-square contribution (CSC) are introduced. The CSC is constructed by employing a moving window and the effect of the window width on its diagnosis performance is investigated. The failure cases of the RBC, the effectiveness of the proposed CSC, as well as the comparison of these three methods for fault diagnosis are demonstrated by case studies on two numerical examples and a simulated three-tank system.

Variable Elimination-Based Contribution for Accurate Fault Identification

We propose a new fault identification method, which can describe the contribution of each process variable to a detected fault and identify a faulty variable more accurately than conventional methods. In the proposed method, in addition to a fault detection model that describes normal operating condition (NOC), multiple fault identification models that describe the same NOC are also constructed by eliminating one variable from all monitored variables at a time. After a fault is detected with the fault detection model, the fault detection index, e.g. a combined index of the T2 and Q statistics, is calculated by using each of the fault identification models. When the faulty variable is eliminated, the index does not change before and after the fault occurs. On the other hand, when the normal variable is eliminated, the index is affected by the fault and increases after the fault occurs. Thus, the eliminated variable corresponding to the index that does not increase after the occurrence of the fault is identified as a faulty variable. In the proposed method, the ratio of the average index in NOC to the current index is used as a fault identification index or a contribution. To validate the proposed method, VEC was compared with the reconstruction-based contribution (RBC) through numerical examples. The results have demonstrated that VEC outperformed RBC in fault identification performance both in the linear case and in the nonlinear case.