Content uploaded by Eric Tatara
Author content
All content in this area was uploaded by Eric Tatara on Jan 10, 2014
Content may be subject to copyright.
Monitoring and fault diagnosis of a polymerization reactor by interfacing knowledge-based and multivariate SPM tools
Abstract
An intelligent process monitoring and fault diagnosis environment
is developed by interfacing multivariate statistical process monitoring
(MSPM) techniques and knowledge-based systems (KBS) for monitoring
continuous multivariable process operation. The software is tested by
monitoring the performance of a continuous stirred tank reactor for
polymerization of vinyl acetate. The real-time KBS G2 and its diagnostic
assistant (GDA) tool are integrated with MSPM methods based on canonical
variate state space (CVSS) process models. Fault detection is based on T
2 of state variables and squared prediction errors (SPE)
charts. Contribution plots in G2 are used for determining the process
variables that have contributed to the out-of-control signal indicated
by large T2 and/or SPE values, and GDA is used to diagnose
the source cause of the abnormal process behavior. The MSPM modules
developed in Matlab are linked with G2 and GDA, permitting the use of
MSPM tools for multivariable processes with autocorrelated data. The
presentation will focus on the structure and performance of the
integrated system. On-line SPM of the multivariable polymerization
process is illustrated by simulation studies
... However, even more sophisticated inferential methods have been developed that take into account the special characteristics of different processes. For instance, multiway models [5][6][7][8] have been applied to batch processes, multiblock models [9][10][11] to processes where one can form logical blocks of data, and dynamic models [4,12,13] to autocorrelated continuous processes. In this work we are interested in scrutinizing the process dynamics and then executing multivariate monitoring based on the acknowledgement of the underlying process dynamics. ...
... Traditionally in process monitoring the majority of the models used are based on the i.i.d. principle [4,13], i.e. the measurement vectors are independent and identically distributed or, in other words, not autocorrelated over time. As aforementioned, this is not the case in this particular plant, which gives rise to problems related to sensitivity and robustness. ...
Two novel approaches are presented which take into account the collinearity among variables and the different phenomena occurring at different scales. This is achieved by combining partial least squares (PLS) and multiresolution analysis (MRA). In this work the two novel approaches are interconnected. First, a standard exploratory PLS model is scrutinized with MRA. In this way, different events at different scales and latent variables are recognized. In this case, especially periodic seasonal fluctuations and long-term drifting introduce problems. These low-frequency variations mask and interfere with the detection of small and moderate-level transient phenomena. As a result, the confidence limits become too wide. This relatively common problem caused by autocorrelated measurements can be avoided by detrending. In practice, this is realized by using fixed-size moving windows and by detrending these windows. Based on the MRA of the standard model, the second PLS model for process monitoring is constructed based on the filtered measurements. This filtering is done by removing the low-frequency scales representing low-frequency components, such as seasonal fluctuations and other long-term variations, prior to standard PLS modeling. For these particular data the results are shown to be superior compared to a conventional PLS model based on the non-filtered measurements. Often, model updating is necessary owing to non-stationary characteristics of the process and variables. As a big advantage, this new approach seems to remove any further need for model updating, at least in this particular case. This is because the presented approach removes low-frequency fluctuations and results in a more stationary filtered data set that is more suitable for monitoring. Copyright © 2000 John Wiley & Sons, Ltd.
... A biological batch process for the treatment of wastewater has been used to develop and test the supervision method. Multivariate Statistic Process Control (MSPC) methods have shown to be effective in detecting and diagnosing events that cause a significant change in the dynamic correlation structure among the process variables [3] some examples are: polymerization reactor process [12] , pharmaceutical process [7], the elaboration at industrial scale of the polymer polypropylene oxide [20], WasteWater Treatment Plant [6] ...
The paper focuses on the development of a classification strategy to identify critic situation in batch process control. Data acquired from a batch execution is reduced by means of multiway principal component analysis in order to be assessed according to the statistical model of the process. Multiple situations have been categorized by a classification algorithm applied to the principal components in order to identify misbehaviour causes.
... A biological batch process for the treatment of wastewater has been used to develop and test the supervision method. Multivariate Statistic Process Control (MSPC) methods have shown to be effective in detecting and diagnosing events that cause a significant change in the dynamic correlation structure among the process variables [3] some examples are: polymerization reactor process [12] , pharmaceutical process [7], the elaboration at industrial scale of the polymer polypropylene oxide [20], WasteWater Treatment Plant [6] ...
The paper focuses on the development of a classification strategy to identify critic situation in batch process control. Data acquired from a batch execution is reduced by means of multiway principal component analysisin order to be assessed according to the statistical model of the process. Multiple situations have beencategorized by a classification algorithm applied to the principal components in order to identify is behaviour causes.
A combination of Multivariate Statistical Process Control (MSPC) and an automatic classification algorithm is applied to monitor a Waste Water Treatment Plant (WWTP). The goal of this work is to evaluate the capabilities of these techniques for assessing the actual state of a WWTP. The research was performed in a pilot WWTP operating with a Sequencing Batch Reactor (SBR). The results obtained refer to the dependence of process behavior with environmental conditions and the identification of specific abnormal operating conditions. It turned out that the combination of tolls yields better classifications compared with those obtained by using methods based on Partial Least Squares.
This paper proposes a model-based detection and isolation (FDI) system based on nonlinear state estimation that can be applied to nonlinear systems. The proposed FDI system uses an extended Kalman filter (EKF), in which conditions based on high filtering are defined to best serve the FDI objectives. A better understanding of the residual trends, calculated from the difference between measurements and the EKF estimates, can be obtained when a fault occurs by developing a model that is able to predict the behavior of the residuals. This model is utilized as the basis for detection and isolation of single and multiple faults. Comparisons with data driven techniques, specifically principal component analysis (PCA) and Kernel PCA, show superior isolation results, having the advantage of distinguishing single and multiple faults from a diverse array of possible faults, a common occurrence in complex processes. The proposed approach is validated using an experimental air heater.
The paper deals with the use of coupled bond graph as an integrated decision tool for health monitoring of chemical reactors. A bond graph model based diagnostic strategy is adopted to detect and isolate kinetic and thermodynamic drift of chemical reactors due to the appearance of secondary reactions. Furthermore, the causal and structural properties of the bond graph model are used not only for system dynamics simulation but also to design Fault Detection and Isolation (FDI) algorithms (i.e. the generation of formal fault indicators) for online supervision of faults affecting sensors, actuators, physico-chemical components and phenomena.
The degree of mixing in polymerization reactions can be influenced by various factors that can also affect the reactor performance. For this reason, a detailed micromixing model was implemented to study the effects of micromixing on the dynamic behavior of continuous free-radical solution polymerization tank reactors. The reactor model was used to perform the bifurcation analysis of the reacting system, paying special attention to the effect of micromixing parameters on the reactor behavior. The bifurcation study showed that multiple steady-states and periodic oscillations can be observed under partially segregated micromixing conditions. Moreover, the micromixing model was able to describe the dynamic responses presented by perfectly mixed and completely segregated reactors. These results indicate that this class of reactors can exhibit more complex dynamic behavior than shown until now.
Industrial continuous processes may have a large number of process variables and are usually operated for extended periods at fixed operating points under closed-loop control, yielding process measurements that are autocorrelated, cross-correlated, and collinear. A statistical process monitoring (SPM) method based on multivariate statistics and system theory is introduced to monitor the variability of such processes. The statistical model that describes the in-control variability is based on a canonical-variate (CV) state-space model that is an equivalent representation of a vector autoregressive moving-average time-series model. The CV state variables obtained from the state-space model are linear combinations of the past process measurements that explain the variability of the future measurements the most. Because of this distinctive feature, the CV state variables are regarded as the principal dynamic directions A T2 statistic based on the CV state variables is used for developing an SPM procedure. Simple examples based on simulated data and an experimental application based on a high-temperature short-time milk pasteurization process illustrate advantages of the proposed SPM method.
Statistical process control methods for monitoring processes with multivariate measurements in both the product quality variable space and the process variable space are considered. Traditional multivariate control charts based on X2 and T2 statistics are shown to be very effective for detecting events when the multivariate space is not too large or ill-conditioned. Methods for detecting the variable(s) contributing to the out-of-control signal of the multivariate chart are suggested. Newer approaches based on principal component analysis and partial least squares are able to handle large ill-conditioned measurement space; they also provide diagnostics which can point to possible assignable causes for the event. The me hods are illustrated on a simulated process of a high pressure low density polyethylene reactor, and examples of their application to a variety of industrial processes are referenced.
Very general reduced order filtering and modeling problems are phased in terms of choosing a state based upon past information to optimally predict the future as measured by a quadratic prediction error criterion. The canonical variate method is extended to approximately solve this problem and give a near optimal reduced-order state space model. The approach is related to the Hankel norm approximation method. The central step in the computation involves a singular value decomposition which is numerically very accurate and stable. An application to reduced-order modeling of transfer functions for stream flow dynamics is given.