Publications Scientifiques

Permanent URI for this communityhttps://dspace.univ-boumerdes.dz/handle/123456789/10

Browse

Author: search.filters.author.Kouadri, A.

Search Results

Now showing 1 - 9 of 9
  • No Thumbnail Available
    Item
    New reduced kernel PCA for fault detection and diagnosis in cement rotary kiln
    (Elsevier, 2020) Bencheikh, Fares; Harkat, M. F.; Kouadri, A.; Bensmail, A.
    Fault detection and diagnosis (FDD) based on data-driven techniques play a crucial role in industrial process monitoring. It intends to promptly detect and identify abnormalities and enhance the reliability and safety of the processes. Kernel Principal Component Analysis (KPCA) is a powerful FDD based data-driven method. It has gained much interest due to its ability in monitoring nonlinear systems. However, KPCA suffers from high computing time and large storage space when a large-sized training dataset is used. So, extracting and selecting the more relevant observations could provide a good solution to high computation time and memory re- quirements costs. In this paper, a new Reduced KPCA (RKPCA) approach is developed to address that issue. It aims to preserve one representative observation for each similar and selected Euclidean distance between training samples. Afterwards, the obtained reduced training dataset is used to build a KPCA model for FDD purposes. The developed RKPCA scheme is tested and evaluated across a numerical example and an actual involuntary process fault and various simulated sensor faults in a cement plant. The obtained results show high monitoring perfor- mance with highest robustness to false alarms and maximum fault detection sensitivity compared to conventional PCA, KPCA and other well-established RKPCA techniques. Furthermore, the unified contribution plot method demonstrates superior potentials in identifying faulty variables.
  • No Thumbnail Available
    Item
    Machine Learning-Based Reduced Kernel PCA Model for Nonlinear Chemical Process Monitoring
    (Springer, 2020) Harkat, M.-F.; Kouadri, A.; Fezai, R.; Mansouri, M.; Nounou, H.; Nounou, M.
    Principal component analysis (PCA) is a popular tool for linear dimensionality reduction and fault detection. Kernel PCA (KPCA) is the nonlinear form of the PCA, which better exploits a complicated spatial structure of high-dimensional features, where a kernel function implicitly defines a nonlinear transformation into a feature space wherein standard PCA is performed. Despite its success and flexibility, conventional KPCA might not perform properly because the use of KPCA for a large-sized training dataset imposes a high computational load and a significant storage memory space since the required elements used for modelling have to be saved and used for monitoring, as well. To address this problem, a reduced KPCA (RKPCA) for fault detection of chemical processes is developed. RKPCA is a novel machine learning tool which merges dimensionality reduction, supervised learning as well as kernel selection. This novel method is used to reduce the size of recorded measurements while maintaining the most relevant data features. The removed observations, including redundant samples that are linearly correlated in the collected measurements, are described by only one sample. The obtained uncorrelated observations via PCA technique are then employed to identify the reduced KPCA model by which Hotelling T2 and squared predictive error or Q statistics are extracted for detection purposes. Besides, their combination is also used as a detection index. The performance of the proposed process monitoring technique is illustrated through its application to Tennessee Eastman process. The obtained results demonstrate the effectiveness of the developed RKPCA technique in detecting various faults with remarkably reduced computation time and memory storage space
  • No Thumbnail Available
    Item
    Reliable fault detection and diagnosis of large-scale nonlinear uncertain systems using interval reduced kernel PLS
    (Institute of Electrical and Electronics Engineers, 2020) Fezai, R.; Abodayeh, K.; Mansouri, M.; Kouadri, A.; Harkat, M.-F.; Nounou, H.; Nounou, M.
    Kernel partial least squares (KPLS) models are widely used as nonlinear data-driven methods for faults detection (FD) in industrial processes. However, KPLS models lead to irrelevant performance over long operation periods due to process parameters changes, errors and uncertainties associated with measurements. Therefore, in this paper, two different interval reduced KPLS (IRKPLS) models are developed for monitoring large scale nonlinear uncertain systems. The proposed IRKPLS models present an interval versions of the classical KPLS model. The two proposed IRKPLS models are based on the Euclidean distance between interval-valued observations as a dissimilarity metric to keep only the more relevant and informative samples. The first proposed IRKPLS technique uses the centers and ranges of intervals to estimate the interval model, while the second one is based on the upper and lower bounds of intervals for model identification. These obtained models are used to evaluate the monitored interval residuals. The aforementioned interval residuals are fed to the generalized likelihood ratio test (GLRT) chart to detect the faults. In addition to considering the uncertainties in the input-output systems, the new IRKPLS-based GLRT techniques aim to decrease the execution time when ensuring the fault detection performance. The developed IRKPLS-based GLRT approaches are evaluated across various faults of the well-known Tennessee Eastman (TE) process. The performance of the proposed IRKPLS-based GLRT methods is evaluated in terms of missed detection rate, false alarms rate, and execution time. The obtained results demonstrate the efficiency of the proposed approaches, compared with the classical interval KPLS
  • No Thumbnail Available
    Item
    False alarms rate reduction using filtered monitoring indices
    (UMBB, 2017) Ammiche, Mustapha; Kouadri, A.
    False alarms are the major problem in fault detection when using multivariate statistical process monitoring such as principal component analysis (PCA), they affect the detection accuracy and lead to make wrong decisions about the process operation status. In this work, filtering the monitoring indices is proposed to enhance the detection by reducing the number of false alarms. The filters that were used are: Standard Median Filter (SMF), Improved Median Filter (IMF) and fuzzy logic based filter. Signal to Noise Ratio (SNR), False Alarms Rate (FAR) and the detection time of the fault were used as criteria to compare their performance and their filtering action influence on monitoring. The algorithms were applied to cement rotary kiln data; real data, to remove spikes and outliers on the monitoring indices of PCA, and then, the filtered signals were used to supervise the system. The results, in which the fuzzy logic based filter showed a satisfactory performance, are presented and discussed
  • No Thumbnail Available
    Item
    Prediction of boiler output variables through the PLS linear regression technique
    (2011) Kouadri, A.; Zelmat, M.; Albarbar, A.
  • No Thumbnail Available
    Item
  • No Thumbnail Available
    Item
    A statistical-based approach for fault detection in a three tank system
    (Taylor & Francis, 2013) Kouadri, A.; Namoun, A.; Zelmat, M.; Aitouche, Moh-Amokrane
    Fault detection in stochastic dynamical systems is usually carried out by the generation of residuals directly reflecting the magnitude of the faults. For this purpose, faults indicator is used to evaluate possible deviations from the normal operating conditions and the measurements of the system. This evaluation is often very difficult to implement in the multi-faults case. This article aims to demonstrate the efficiency of the coefficient of variation (CV) in detecting single and multi-faults in a multivariable laboratory three tank system DTS-200. The performance of the detection algorithm is based on the computation of the confidence intervals (CIs) which provide an estimate of the amount of error in the considered data and characterise the precision of the computed statistical estimates. The data variability may result from random measurement errors caused by the system parameters uncertainties, internal and external noises, and measuring instrument, which are not usually accurate. The CIs make the CV less sensitive to parameter uncertainties and to measure noises. The robustness and accuracy of the CV are shown in a healthy mode and various faulty situations in an entirely uncertain environment
  • No Thumbnail Available
    Item
    A radial basis function neural network optimized through modified DIRECT algorithm based-model for a three interconnected water tank
    (Taylor & Francis, 2010) Kouadri, A.; Chiter, L.; Zelmat, M.
    In many physical systems, it is difficult to obtain a model structure that is highly nonlinear and complex. However, models are usually linear, but not suitable in such form to model processes because they contain a significant number of simplifying hypotheses which are insufficient for the design of reliable controllers. The absence of robustness with respect to system parameters does not ensure the performance specifications of the control system knowing that the nominal parametric state rarely corresponds to the real one. For these raisons, it is beneficial to use a specific technique to characterize accurately system dynamics in an entirely uncertain environment. In this work, we present an approach to approximate and validate over a large operating range the dynamic behaviour of a Three Tank System benchmark based on a radial basis function neural network (RBFNN). The proposed RBFNN is applied to solve the parametric-identification problems for nonlinear and complex system by using a modified DIRECT algorithm to search the network parameters. The learning algorithm is developed by combining the DIRECT algorithm and a linear regression for fast convergence. Different experimental results have been performed to show the effectiveness of the RBFNN model to emulate the dynamic behaviour of the nonlinear and complex system under different situations
  • No Thumbnail Available
    Item
    Boundary effects on confined swirling flows with vortex breakdown
    (2006) Saci, R.; Kouadri, A.; Saadi, S.
    Confined steady swirling flows, driven by the end disks of a cylindrical/truncated conical enclosure have been numerically studied. Particular attention is focused on combined kinematics and geometric conditions of generation and control of the vortex breakdown phenomenon. First, the basic steady flow topology in a truncated conical cavity is described, which is shown to depend strongly on the direction as well as the rate of rotation of the end disks. For a set of governing flow parameters, the computations revealed the occurrence of bubble-like reverse flows, characterised by onaxis stagnation points. The present work, explores means of controlling the evolution of this physical phenomenon, by modifying the boundary conditions upstream the vortex breakdown. These means are found to either suppress or enhance the occurrence and size of the bubbles