Multiple faults detection and isolation approach applied to a kraft recovery boiler

TAPPI Journal ◽  
2014 ◽  
Vol 13 (1) ◽  
pp. 33-41
Author(s):  
YVON THARRAULT ◽  
MOULOUD AMAZOUZ
Keyword(s):  
Fault Detection ◽  
Early Detection ◽  
False Alarm ◽  
False Alarm Rate ◽  
Principal Component ◽  
Fault Isolation ◽  
Fault Detection And Isolation ◽  
Detection Scheme ◽  
Multiple Faults ◽  
Recovery Boiler

Recovery boilers play a key role in chemical pulp mills. Early detection of defects, such as water leaks, in a recovery boiler is critical to the prevention of explosions, which can occur when water reaches the molten smelt bed of the boiler. Early detection is difficult to achieve because of the complexity and the multitude of recovery boiler operating parameters. Multiple faults can occur in multiple components of the boiler simultaneously, and an efficient and robust fault isolation method is needed. In this paper, we present a new fault detection and isolation scheme for multiple faults. The proposed approach is based on principal component analysis (PCA), a popular fault detection technique. For fault detection, the Mahalanobis distance with an exponentially weighted moving average filter to reduce the false alarm rate is used. This filter is used to adapt the sensitivity of the fault detection scheme versus false alarm rate. For fault isolation, the reconstruction-based contribution is used. To avoid a combinatorial excess of faulty scenarios related to multiple faults, an iterative approach is used. This new method was validated using real data from a pulp and paper mill in Canada. The results demonstrate that the proposed method can effectively detect sensor faults and water leakage.

scholarly journals Application of a Fault Detection and Isolation System on a Rotary Machine

2013 ◽  
Vol 2013 ◽  
pp. 1-11 ◽  
Author(s):  
Silvia M. Zanoli ◽  
Giacomo Astolfi
Keyword(s):  
Fault Detection ◽  
Centrifugal Compressor ◽  
Principal Component ◽  
Fault Isolation ◽  
Fault Detection And Isolation ◽  
Isolation System ◽  
Multiple Faults ◽  
Rotary Machine ◽  
Model Free ◽  
Refinery Plant

The paper illustrates the design and the implementation of a Fault Detection and Isolation (FDI) system to a rotary machine like a multishaft centrifugal compressor. A model-free approach, that is, the Principal Component Analysis (PCA), has been employed to solve the fault detection issue. For the fault isolation purpose structured residuals have been adopted while an adaptive threshold has been designed in order to detect and to isolate the faults. To prove the goodness of the proposed FDI system, historical data of a nitrogen centrifugal compressor employed in a refinery plant are considered. Tests results show that detection and isolation of single as well as multiple faults are successfully achieved.

Image-Based Versus Signal-Based Sensors for Machine Fault Detection and Isolation

2014 ◽  
Author(s):  
Heshan Fernando ◽  
Vedang Chauhan ◽  
Brian Surgenor
Keyword(s):  
Fault Detection ◽  
High Speed ◽  
Fault Isolation ◽  
Video Data ◽  
Fault Detection And Isolation ◽  
Single Camera ◽  
Multiple Faults ◽  
Multiple Sensors ◽  
Machine Fault ◽  
And Performance

This paper presents the results of a comparative study that investigated the use of image-based and signal-based sensors for fault detection and fault isolation of visually-cued faults on an automated assembly machine. The machine assembles 8 mm circular parts, from a bulk-supply, onto continuously moving carriers at a rate of over 100 assemblies per minute. Common faults on the machine include part jams and ejected parts that occur at different locations on the machine. Two sensor systems are installed on the machine for detecting and isolating these faults: an image-based system consisting of a single camera and a signal-based sensor system consisting of multiple greyscale sensors and limit switches. The requirements and performance of both systems are compared for detecting six faults on the assembly machine. It is found that both methods are able to effectively detect the faults but they differ greatly in terms of cost, ease of implementation, detection time and fault isolation capability. The conventional signal-based sensors are low in cost, simple to implement and require little computing power, but the installation is intrusive to the machine and readings from multiple sensors are required for faster fault detection and isolation. The more sophisticated image-based system requires an expensive, high-resolution, high-speed camera and significantly more processing power to detect the same faults; however, the system is not intrusive to the machine, fault isolation becomes a simpler problem with video data, and the single camera is able to detect multiple faults in its field of view.

Fault detection and isolation in nonlinear systems with partial Reduced Kernel Principal Component Analysis method

2016 ◽  
Vol 40 (4) ◽  
pp. 1289-1296 ◽  
Author(s):  
Ines Jaffel ◽  
Okba Taouali ◽  
Mohamed Faouzi Harkat ◽  
Hassani Messaoud
Keyword(s):  
Principal Component Analysis ◽  
Fault Detection ◽  
Principal Component ◽  
Component Analysis ◽  
Fault Isolation ◽  
Fault Detection And Isolation ◽  
Kernel Principal Component Analysis ◽  
Continuous Stirred Tank Reactor ◽  
Analysis Method ◽  
Continuous Stirred Tank

In this article, we suggest an extension of our proposed method in fault detection called Reduced Kernel Principal Component Analysis (RKPCA) (Taouali et al., 2015) to fault isolation. To this end, a set of structured residues is generated by using a partial RKPCA model. Furthermore, each partial RKPCA model was performed on a subset of variables to generate structured residues according to a properly designed incidence matrix. The relevance of the proposed algorithm is revealed on Continuous Stirred Tank Reactor.

scholarly journals Fault Detection and Isolation with Robust Principal Component Analysis

2008 ◽  
Vol 18 (4) ◽  
pp. 429-442 ◽  
Author(s):  
Yvon Tharrault ◽  
Gilles Mourot ◽  
José Ragot ◽  
Didier Maquin
Keyword(s):  
Principal Component Analysis ◽  
Fault Detection ◽  
Covariance Matrix ◽  
Principal Component ◽  
Component Analysis ◽  
Fault Detection And Isolation ◽  
Robust Principal Component Analysis ◽  
Sample Mean ◽  
Simple Estimate ◽  
Multiple Faults

Fault Detection and Isolation with Robust Principal Component AnalysisPrincipal component analysis (PCA) is a powerful fault detection and isolation method. However, the classical PCA, which is based on the estimation of the sample mean and covariance matrix of the data, is very sensitive to outliers in the training data set. Usually robust principal component analysis is applied to remove the effect of outliers on the PCA model. In this paper, a fast two-step algorithm is proposed. First, the objective was to find an accurate estimate of the covariance matrix of the data so that a PCA model might be developed that could then be used for fault detection and isolation. A very simple estimate derived from a one-step weighted variance-covariance estimate is used (Ruiz-Gazen, 1996). This is a "local" matrix of variance which tends to emphasize the contribution of close observations in comparison with distant observations (outliers). Second, structured residuals are used for multiple fault detection and isolation. These structured residuals are based on the reconstruction principle, and the existence condition of such residuals is used to determine the detectable faults and the isolable faults. The proposed scheme avoids the combinatorial explosion of faulty scenarios related to multiple faults to be considered. Then, this procedure for outliers detection and isolation is successfully applied to an example with multiple faults.

Data Segmentation Criteria Assessment for Fault Detection Techniques Based on Principal Component Analysis for Natural Gas Transmission System

2018 ◽  
Author(s):  
Cinthia Audivet ◽  
Horacio Pinzón ◽  
Jesus García ◽  
Marlon Consuegra ◽  
Javier Alexander ◽  
...  
Keyword(s):  
Principal Component Analysis ◽  
Natural Gas ◽  
Fault Detection ◽  
False Alarm ◽  
False Alarm Rate ◽  
Principal Component ◽  
Detection Time ◽  
Case Scenario ◽  
Data Segmentation ◽  
Natural Gas Transmission

Statistical analytics, as a data extraction and fault detection strategy, may incorporate segmentation techniques to overcome its underlying limitations and drawbacks. Merging both techniques shall provide a more robust monitoring structure to address the proper identification of normal and abnormal conditions, to improve the extraction of fundamental correlation among variables, and to improve the separation of both main variation and natural variation (noise) subspaces. This additional feature is key to limit the false alarm rate and to optimize the fault detection time when it is implemented on industrial applications. This paper presents an analysis to determine whether a segmentation approach, as a previous step of detection, enhances the fault detection strategies, specifically the principal component analysis performance. The data segmentation criteria assessed in this study includes two approaches: a) Sources (well) of the transmitted natural gas and b) Promigas’ natural gas pipeline division defined by the Energy and Gas Regulation Commission (CREG in Spanish). The performance assessment of segmentation criteria was carried out evaluating the false alarm rate and detection time when the natural gas transmission network presents faults of different magnitude. The results show that the implementation of a segmentation criteria provides an advantage in terms of the detection time, but it depends of the fault magnitude and the number of clusters. The detection time is improved by 25% in the case scenario I, when transition zones are considered. On the other hand, the detection time is slightly better with less than 10% in the case scenario II, where the segmentation is geographical.

scholarly journals Fault Detection and Isolation in Inertial Measurement Units Based on -CUSUM and Wavelet Packet

2013 ◽  
Vol 2013 ◽  
pp. 1-10 ◽  
Author(s):  
Élcio Jeronimo de Oliveira ◽  
Ijar Milagre da Fonseca ◽  
Hélio Koiti Kuga
Keyword(s):  
Fault Detection ◽  
Wavelet Packet ◽  
Median Filter ◽  
Principal Component ◽  
Fault Isolation ◽  
Fault Detection And Isolation ◽  
Measurement Unit ◽  
Short Delay ◽  
Inertial Measurement ◽  
Low Level

The aim of this paper is to present a fault detection algorithm (FDI) based on signal processing techniques developed for an inertial measurement unit (IMU) with minimal redundancy of fiber optic gyros. In this work the recursive median filter is applied in order to remove impulses (outliers) arising from data acquisition process and parity vector operations, improving the fault detection and isolation performance. The FDI algorithm is divided into two blocks: fault detection (FD) and fault isolation (FI). The FD part of the algorithm is used to guarantee the reliability of the isolation part and is based on parity vector analysis using -CUSUM algorithm. The FI part is performed using parity space projection of the energy subbands obtained from wavelet packet decomposition. This projection is an extension of clustering analysis based on singular value decomposition (SVD) and principal component analysis (PCA). The results of the FD and FI algorithms have shown the effectiveness of the proposed method, in which the FD algorithm is capable of indicating the low-level step bias fault with short delay and a high index of correct decisions of the FI algorithm also with low-level step bias fault.

Isolating observer for simultaneous structural-actuator fault detection

2018 ◽  
Vol 6 (3) ◽  
pp. 118-133
Author(s):  
Hamed Pourazad ◽  
Javad Askari ◽  
Saeed Hosseinnia
Keyword(s):  
Fault Detection ◽  
Detection System ◽  
Design Procedure ◽  
Fault Isolation ◽  
Unmanned Aircraft ◽  
Fault Detection And Isolation ◽  
Simultaneous Occurrence ◽  
Detection Scheme ◽  
Content Type ◽  
Commercial Applications

Purpose Increasing commercial applications for small unmanned aircraft create growing challenges in providing safe flight conditions. The conventional measures to detect icing are either expensive, energy consuming or heavy. The purpose of this paper is to develop a fault identification and isolation scheme using unknown input observers to detect and isolate actuator and structural faults in simultaneous occurrence. Design/methodology/approach The fault detection scheme is based on a deviation in system parameters due to icing and lock-in-place (LIP), two faults from different categories with similar indications that require different reconfiguration actions. The obtained residual signals are selected to be triggered by desired faults, while insensitive to others. Findings The proposed observer is sensitive to both actuator and structural faults, and distinguishes simultaneous occurrences by insensitivity to LIP in selected residue signals. Simulation results confirm the success of the proposed system in the presence of uncertainty and disturbance. Research limitations/implications The fault detection and isolation scheme proposed here is based on the linear model of a winged aircraft, the Aerosonde. Moreover, the faults are applied to rudder and aileron in simulations, but the design procedure for other models is provided. The designed scheme could be further implemented on a non-linear aircraft model. Practical implications Applying the proposed icing detection scheme increases detection system reliability, since fault isolation enables timely reconfiguration schemes. Originality/value The observers proposed in previous papers detected icing fault but were not insensitive to actuator faults.

scholarly journals Estimation and Fault Diagnosis of Lithium-Ion Batteries: A Fractional-Order System Approach

2018 ◽  
Vol 2018 ◽  
pp. 1-12 ◽  
Author(s):  
Shulan Kong ◽  
Mehrdad Saif ◽  
Guozeng Cui
Keyword(s):  
Fault Diagnosis ◽  
Fault Detection ◽  
Fractional Order ◽  
Sliding Mode ◽  
Estimation Error ◽  
Lithium Ion ◽  
Fault Isolation ◽  
Sliding Mode Observer ◽  
Fault Detection And Isolation ◽  
Order System

This study investigates estimation and fault diagnosis of fractional-order Lithium-ion battery system. Two simple and common types of observers are designed to address the design of fault diagnosis and estimation for the fractional-order systems. Fractional-order Luenberger observers are employed to generate residuals which are then used to investigate the feasibility of model based fault detection and isolation. Once a fault is detected and isolated, a fractional-order sliding mode observer is constructed to provide an estimate of the isolated fault. The paper presents some theoretical results for designing stable observers and fault estimators. In particular, the notion of stability in the sense of Mittag-Leffler is first introduced to discuss the state estimation error dynamics. Overall, the design of the Luenberger observer as well as the sliding mode observer can accomplish fault detection, fault isolation, and estimation. The effectiveness of the proposed strategy on a three-cell battery string system is demonstrated.

Optimal Sensor Placement for Damage Detection in Complex Structures

2009 ◽  
Author(s):  
Sunilkumar Soni ◽  
Santanu Das ◽  
Aditi Chattopadhyay
Keyword(s):  
False Alarm ◽  
Damage Detection ◽  
False Alarm Rate ◽  
Structural Component ◽  
Detection Rate ◽  
Sensor Placement ◽  
False Alarms ◽  
Optimal Sensor Placement ◽  
Detection Scheme ◽  
Sensing Radius

An optimal sensor placement methodology is proposed based on detection theory framework to maximize the detection rate and minimize the false alarm rate. Minimizing the false alarm rate for a given detection rate plays an important role in improving the efficiency of a Structural Health Monitoring (SHM) system as it reduces the number of false alarms. The placement technique is such that the sensor features are as directly correlated and as sensitive to damage as possible. The technique accounts for a number of factors, like actuation frequency and strength, minimum damage size, damage detection scheme, material damping, signal to noise ratio (SNR) and sensing radius. These factors are not independent and affect each other. Optimal sensor placement is done in two steps. First, a sensing radius, which can capture any detectable change caused by a perturbation and above a certain threshold, is calculated. This threshold value is based on Neyman-Pearson detector that maximizes the detection rate for a fixed false alarm rate. To avoid sensor redundancy, a criterion to minimize sensing region overlaps of neighboring sensors is defined. Based on the sensing region and the minimum overlap concept, number of sensors needed on a structural component is calculated. In the second step, a damage distribution pattern, known as probability of failure distribute, is calculated for a structural component using finite element analysis. This failure distribution helps in selecting the most sensitive sensors, thereby removing those making remote contributions to the overall detection scheme.

Sign in / Sign up

Export Citation Format

Share Document