![]() ![]() As step change occurs, differencing shows a significant shift in process monitoring, but subsequent faulty observations tend to be under control as they are in relative control of each other. This problem arises when the identification of certain forms of fault, such as step change, emerges. But this method causes a major problem in process monitoring. Ĭoncerning the characteristic of autocorrelation, the first choice could be minimizing the effect of autocorrelation by differencing and then using static PCA. Since a formula is used to estimate the new observation, which has the least similarity with the measurement time parameters of the new observation. For example, if a static model is implemented on non-stationery data, the model structure is inefficient to estimate the new observation because the mean and variance criteria change over time. Therefore, although it is true that quality control charts based on conventional PCA are capable of managing the high dimensional process and suitable for stationary one, the findings suggested that linear and non-adaptive methods have high false alarm rates because of the incompatibility of this approach with auto-correlated and non-stationary process data. However, in addition to their high dimensionality, data reported to solve sophisticated industrial, health care, IT, or economic problems have characteristics such as autocorrelation and non-stationary nature. One of the data-driven multivariate statistical tools is the quality control charts by the principal component analysis approach (PCA) to detect an abnormal behavior. Īs a consequence, a multivariate statistical approach is needed when the number of variables or dimensions of an industrial problem is beyond one value. Data-driven approaches are chosen upon the availability of product or system data, but the system model is not. Hence, data-driven methods can be used for large-scale and complex systems, which are inexpensive as well. To detect faults, the data-driven methods use products’ life-cycle data, which they are not dependent on the first-principles. When the detailed mathematical model cannot be reached, and when the number of inputs, outputs, and states of a system is logically limited, the knowledge-based method will yield the best results. The failure cases and engineers’ experience are used to formulate the rules. , the knowledge-based methods are mostly rule-based expert systems. The analytical model cannot be applied for large-scale and complex systems. ![]() The analytical approach utilizes the first principles to construct mathematical models of the system. The results demonstrate that the proposed approach has detected a real fault successfully.Ĭlassification of fault detection and diagnosis method. The empirical application of the proposed approach has been implemented on a turbine exit temperature (TET). The findings suggest that the proposed approach is capable of detecting various forms of faults and comparing attempts to improve the detection of fault indicators with other approaches. The proposed approach has been tested on the Tennessee Eastman Process (TEP). This technique utilizes DPCA property to decrease the effect of autocorrelation and adaptive behavior of MWPCA to control non-stationary characteristics. A new PCA monitoring method is proposed in this study, which can simultaneously reduce the impact of high-dimensionality, non-stationary, and autocorrelation properties. ![]() But, using the techniques listed without considering all aspects of the process data increases fault detection indicators such as false alarm rate (FAR), delay time detection (DTD), and confuses the operator or causes adverse consequences. To date, approaches such as the recursive PCA (RPCA) model and the moving-window PCA (MWPCA) model have been proposed when data is high-dimensional and non-stationary or dynamic PCA (DPCA) model and its extension have been suggested for autocorrelation data. An efficient fault detection technique is an approach that is robust against data training, sensitive to all the feasible faults of the process, and agile to the detection of the faults. Industrial processing data involves complexities such as high dimensionality, auto-correlation, and non-stationary which may occur simultaneously. The control charts with the Principal Component Analysis (PCA) approach and its extension are among the data-driven methods for process monitoring and the detection of faults. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |