Search

Fault diagnosis(FD) is vital for monitoring of industrial processes. Actually, both the multiple fault variables and minor faults are inclined to bring wrong diagnosis results. In this paper, a fault variable identification method is proposed for multivariate/minor fault diagnosis. The deviation factor is adopted as the characteristic of the sample, and Bayesian decision theory is adopted to calculate the possibility of the variable being faulty, then multi-dimensional reconstruction-based contribution (MRBC) is used to determine the fault source variables. This method not only improves the diagnosis rate for multiple/minor faults with less computation time, but also can indicate severity level of the different fault variables according to the fault occurrence probability. A Numerical example and Tennessee Eastman process are given to show the efficiency of the proposed method. • A fault variable identification method is proposed for multivariate/minor fault. • The deviation factor, the Bayesian decision theory and MRBC is used for FD. • The proposed method can indicate severity level of the different fault variables. • The proposed method will improve the diagnosis rate with less computation time. [ABSTRACT FROM AUTHOR]

Traditional data-driven methods generally suppose the training dataset is not corrupted by outliers. However, outliers are inevitable in the real industrial processes even with a relatively high ratio, which degrades the accuracy of data-based models. For multimode process monitoring, outliers may deteriorate the accuracy of both mode identification and fault detection. However, the existing robust methods can hardly deal with a large percentage of outliers, i.e., dirty data in neither single nor multimode processes. In this paper, a robust multimode process monitoring scheme is developed by alleviating the negative effect of dirty data. A difference-based decomposition of matrix (DDM) algorithm is first proposed to divide the data into a basic subpart and a non-basic subpart. The optimization function of the proposed DDM algorithm is solved by the alternating direction method of multipliers (ADMM). In the off-line procedure, an iterative decomposition strategy is designed based on the proposed DDM approach to identify dirty data and obtain the clean dataset of each mode. In the on-line procedure, an on-line sample identification strategy is developed by the type indicator derived from the DDM algorithm to determine whether the current sample belongs to a mode, the dirty data, or the fault. A numerical example and an industrial-scale multiphase flow facility indicate the proposed method can improve the accuracy of both mode identification and fault detection for multimode processes even under dirty data. • The proposed method monitors multimode processes corrupted with dirty data. • The proposed DDM algorithm selects non-similar data from a certain class of data. • The method detects dirty data and identifies modes concurrently without mode numbers. • Numerical case and multiphase flow process proved the effectiveness of our proposal. [ABSTRACT FROM AUTHOR]

Many continuous industrial processes operate in different steady states with different grades or products. The switching between two steady states is called transition. Transition consists of a series of operation changes that should be carried out in proper order, within certain magnitudes and time region. Since faulty operation may lead to increase in inferior products or even hazard events, monitoring of the transition is desired. In this work, a transition identification and monitoring scheme is proposed based on slow feature analysis. Two monitoring statistics which represent the location of the trajectory and the speed of transition are proposed. Besides, operating faults are generated based on the guidewords of hazard and operability analysis (HAZOP). Using a numerical case and the mode 4-to-2 transition of the Tennessee-Eastman process in which catastrophic failures exist, the effectiveness of the proposed method is validated. In addition to missed detection rate and false alarm rate, two performance indexes known as detection time (DT) and rescue time (RT) are introduced. The advantages of proposed method are benchmarked against the stage-based sub principle component analysis(sub-PCA) and the global preserving statistics slow feature analysis(GSSFA). • A transition identification and monitoring scheme is proposed based on SFA. • The trajectory-based transion monitoring method is presented with two statistics. • Operating faults for transition switching are generated based on HAZOP. • Two performance indexes (DT and RT) are introduced for the catastrophic failures. [ABSTRACT FROM AUTHOR]

The performance of model-based control systems depends a lot on the process model quality, hence the process model-plant mismatch is an important factor degrading the control performance. In this paper, a new methodology based on a process model evaluation index is proposed for detecting process model mismatch in closed-loop control systems. The proposed index is the ratio between the variance of the disturbance innovation and that of the model quality variable. The disturbance innovations are estimated from the routine operation data by an orthogonal projection method. The model quality variable can be obtained using the closed-loop data and the disturbance model estimated by adaptive Least absolute shrinkage and selection operator (Lasso) method. When the order of the disturbance model is less than 2 or the process time delay is large enough, no external perturbations are required. Besides, the proposed index is independent of the controller tuning and insensitive to the changes in disturbance model, which indicates that the proposed method can isolate the process model-plant mismatch from other factors affecting the overall control performance. Three systems with proportional integral (PI) controller, linear quadratic (LQ) controller and unconstrained model predictive control (MPC) respectively are presented as examples to verify the effectiveness of the proposed technique. Besides, Tennessee Eastman process shows the proposed method is able to detect process model mismatch of nonlinear systems. [ABSTRACT FROM AUTHOR]

Abstract: In semiconductor manufacturing processes, mixed-products are usually fabricated on the same set of process tool with different recipes. Run-to-run controllers which based on the exponential weighted moving average (EWMA) statistic are probably the most frequently used in industry for the quality control of certain semiconductor manufacturing process steps. However, for mixed-product drifted process, if the break length of a product is large, then the process output at the beginning runs of each cycle will far deviate from the target value which will lead to a possible high rework rate and lots of waste wafers. Therefore, this study aims to develop a new approach named cycle forecasting EWMA (CF-EWMA) approach to deal with the problem of large deviations in the first few runs of each cycle. Furthermore, a common fault, i.e., the step fault, is also considered in this paper, and fault tolerant cycle forecasting EWMA (FTCF-EWMA) approach is proposed. Simulation study shows that the proposed approaches are effective. [Copyright &y& Elsevier]

Abstract: Semiconductor manufacturing industry has elevated cost in productions. Improvement of production efficiency is always an important goal for manufacturers. Run-to-run control has been widely used in batch manufacturing processes to reduce variations. Threaded exponentially weighted moving average (threaded-EWMA) run-to-run control is an important and stable control scheme. In this paper, we study the drifted process with mixed products are manufactured in cycles on the same tool, and find that the process outputs will be off target greatly at the beginning runs of cycle 1, 2,… if the product has a long break length. In order to reduce a possible high rework rate, a threaded double EWMA (thread-dEWMA) controller is used to handle the disturbance as well as the drift. By analysis of the system output, a drift-compensatory approach is proposed to eliminate these large deviations. In order to enhance the system performance, the well known “trade-off” solution is adopted to choose the optimal discount factors. Furthermore, how to deal with process fault is also considered in this paper. Two kinds of fault – the step fault and the ramp fault are discussed for fault tolerant approach which can reduce the large deviations of process output from the specification. Simulation study showed that the proposed approaches are effective. [Copyright &y& Elsevier]

Abstract: Run-to-run control has been widely used in batch manufacturing processes to reduce variations. However, in batch processes, many different products are fabricated on the same set of process tool with different recipes. Two intuitive ways of defining a control scheme for such a mixed production mode are (i) each run of different products is used to estimate a common tool disturbance parameter, i.e., a “tool-based” approach, (ii) only a single disturbance parameter that describe the combined effect of both tool and product is estimated by results of runs of a particular product on a specific tool, i.e., a “product-based” approach. In this study, a model two-product plant was developed to investigate the “tool-based” and “product-based” approaches. The closed-loop responses are derived analytically and control performances are evaluated. We found that a “tool-based” approach is unstable when the plant is non-stationary and the plant-model mismatches are different for different products. A “product-based” control is stable but its performance will be inferior to single product control when the drift is significant. While the controller for frequent products can be tuned in a similar manner as in single product control, a more active controller should be used for the infrequent products which experience a larger drift between runs. The results were substantiated for a larger system with multiple products, multiple plants and random production schedule. [Copyright &y& Elsevier]