Fault Detection and Operation Mode Identification Based on Pattern Classification with Variable Selection

Abstract

A novel method is proposed for fault detection and operation mode identification in processes with multimode operations. The proposed method employs the support vector machine as a classification tool together with an entropy-based variable selection method to deal with normal data clusters corresponding to multiple operational modes and abnormal data corresponding to faults. The use of the classification method in fault detection and operation mode identification allows us to build decision boundaries among the data clusters without the assumption of normal distribution. In addition, selection of variables by minimizing the total entropy of training data ensures superior generalization performance in classification. The performance of the proposed method is compared with that of the traditional PCA-based fault detection method using test data where class information is already known. While the existing method has produced considerable error rate (1.9% by T 2 chart and 24.2% by Q chart) in detecting faults, the proposed method has shown no error in this example in either fault detection or operation mode identification. Despite these outstanding results, it should be noted that the performance of the proposed method depends critically on the quality of the training data.

Variable selection method for fault isolation using least absolute shrinkage and selection operator (LASSO)

“…Owing to the similarity between Eqs. (14) and (17), such discussions are also valid for the problem of multivariate fault isolation. The details are as follows.…”

“…When new process measurements indicate an abnormality, the causes of the fault can be identified by determining the degree of similarity with the known event data. Different types of classification methods, such as those involving support vector machines, Fisher discriminant analysis, fuzzy logic, k-means clustering, and fault tree analysis, have all been used as diagnosis tools, e.g., in [1,[13][14][15]. However, in practice it is usually difficult to acquire sufficient historical fault data, which serve as a basis for fault classification.…”

Multimode Monitoring of Oxy-Gas Combustion Through Flame Imaging, Principal Component Analysis, and Kernel Support Vector Machine

“…As one of the traditional classifiers, a support vector machine (SVM) model determines the optimal separation hyperplane that can divide the training set into two classes and has widely been used in many applications in the field of mode identification (Chu et al, 2004;Gunn, 1998). For non-linear mapping data, the kernel function, which represents an inner product between samples in a high-dimensional space, is introduced in the SVM model, so called KSVM which is capable of transferring the non-linear data into a separable space.…”

“…Finally, local-models identify the operation condition by using clustering methods and detect the state of the corresponding condition (Tong and Yan, 2013). For instance, Chu et al (2004) proposed a strategy which employed the support vector machine (SVM) and an entropy-based variable selection method for the fault detection and operation mode identification in processes with multimode operations. Zhu et al (2012) applied k-independent component analysis-principal component analysis (k-ICA-PCA) in the process pattern construction and multimode monitoring.…”

Classification of Household Appliance Operation Cycles: A Case-Study Approach

“…These research works demonstrate that multiple appliances can be distinguished based on certain classification algorithms. It is to be noted that only a handful of research work [8] has been done to study the classification of multiple operation modes of appliances. Some of the household appliances, such as a lamp or television, have single or limited operation modes, while others such as a refrigerator or clothes washer, have multiple operation modes due to their settings.…”