Search

Supervised learning for chemical process fault classifications need for a large amount of sampling data with fault labels which significantly consume expert manpower in general. It is seen that active learning is becoming an effective solution for this problem. To improve the quality of sampling data used for expert labeling, this paper proposes sampling evaluation metrics combing information entropy and similarity for active learning strategies in chemical process fault classifications. Specifically, in response to the insensitive direction of the cosine similarity, the Euclidean distance is merged with evaluate the sampling representativeness. Further, combing with information entropy, a comprehensive evaluation index (E-ECos) is established to select the most valuable samples for expert labeling, adding to the training set of classifiers, by labeling them by experts and adding them to the supervised learning training set of the classifier, the classification learning performance can be effectively improved with a relatively small number of actively labeled samples. On the TE process simulation platform, a deep learning network and an ensemble learning network based classifiers are employed to classify industrial process faults, respectively. Experimental results demonstrate the effectiveness of the proposed method. • This paper proposes a fault classification method with an improved active learning. To improve the quality of sampling data used for expert labeling, this method combing information entropy and similarity for active learning strategies in chemical process fault classifications. • Sample similarities: This paper comprehensively evaluates the similarity of samples by evaluating the cosine similarity and Euclidean distance between samples. On the one hand, Euclidean distance complements the problem that cosine similarity is insensitive to absolute direction. On another hand, the Euclidean distance is based on the absolute value of dimensional features. • E-ECOS: Taking advantage of the different sample selecting evaluation metrics of the two active learning algorithms, we establish a comprehensive evaluation index in terms of sample informativeness and sample representativeness. • Future potentiality: The following research is suggested to combine active learning with auto-machine labeling to further reduce the consumption of manpower for fault sample labeling. [ABSTRACT FROM AUTHOR]

This paper proposes a hierarchical sparse artificial neural network for classifying the faults in dynamic processes base on limited labeled data. The Stacked auto-encoders (SAE) is developed to extract features from different faults. Each neural network in the proposed SAE is given a sparse constraint to learn a Sparse Stacked auto-encoders (SSAE). Then, the Dynamic time window is combined into SSAE to build Dynamic Sparse Stacked auto-encoders (DSSAE). DSSAE model based semi-supervised fault classification scheme is then formulated to classify the dynamic faulty data. Simulation studies on the Tennessee–Eastman (TE) benchmark process evaluate the performance of the developed method, which indicate that the DSSAE method performs better than both SAE and SSAE. [ABSTRACT FROM AUTHOR]

Performance of the most existing fault detection and classification methods can only be guaranteed when each of their own assumptions are met. In other words, a method works well in one condition may not perform well in another. In this paper, a new analytic hierarchy process (AHP) based fuzzy decision fusion system is proposed to tackle the fault classification problem. The AHP approach is introduced to determine the priorities of different classifiers, which are further utilized as the weights in ensemble system. Comparing to conventional equal weighted fusion system, the proposed fuzzy fusion system is able to provide more rational and convincing fault classification result. Effectiveness of the proposed fuzzy fusion system with model evaluation is verified through the Tennessee Eastman (TE) benchmark process. [ABSTRACT FROM AUTHOR]

The fault classification of chemical processes based on supervised learning generally demands for enough labeled fault samples, which represents a big challenge for practical engineering issues. In the industrial field, when capturing an appearance of process abnormities, human operators usually take account of correlations of relevant variables so as to locate process variables associated with the abnormity before further classifying or diagnosing the corresponding faults. In this context, accurate abnormity localizations can be an effective help for classifying the fault type. Nowadays, deep learning based fault classifications have been extensively studied and widely circulated in the literature. However, the existing approaches rarely concern the supporting role of abnormity locations for fault classifications. Motivated by this observation, this paper proposes a multi-task learning (MTL) based convolutional neural network (CNN) model for concurrent abnormity locations and fault classifications for chemical processes. Specifically, fault samples are labeled with both abnormal variable locations and fault categories before constructing a multi-task learning CNN classifier performing an auxiliary task, abnormity locations, and a primary task, fault classifications. In addition, dynamic task weights are assigned to facilitate task contributions. This method can enjoy effectively reducing the necessary amounts of labeled fault samples in achieving satisfactory fault classification performances. Experiments are carried out on the TE process simulation platform and an industrial coal gasification plant along with comparisons with conventional fault classification methods, verifying the effectiveness of the proposed method. • A CNN model based on multi-task learning is proposed to better adapt to fault diagnosis of chemical industrial processes. • An adaptive multitask learning algorithm is proposed to balance the multi-task training process effectively. • Abnormity locations can improve the accuracy of fault classification and the convergence of the overall network training. • Abnormity locations task branches can share tags to the primary task thus reducing the impact of insufficient fault labels. • Future potentiality: Since MTL method doesn't require architectural constraints, it can be applied to other neural networks. [ABSTRACT FROM AUTHOR]

Since industrial process data often presents strong correlations, high complexity, and nonlinear patterns, a proficient deep learning model is required for the fault classification task. Recent researches have shown that deep learning models like stacked autoencoder (SAE) are able to learn deep abstract features from complex process data. Nevertheless, a traditional SAE cannot extract the informative fault-relevant features and data distribution features from industrial process data, which are necessary for effective fault classification in industrial processes. Thus, this study proposes a semi-supervised Inter-Relational Mahalanobis SAE (IRM-SAE) model to learn inter-relational distribution and fault-relevant dynamic features of process data for fault classification. First, the Inter-Relational Mahalanobis loss is introduced into the original objective function of SAE to learn meaningful inter-relational distribution features within the data. Then, an active time frame technique is developed to preprocess the input data to capture the dynamic features of the data. Furthermore, to fully utilize both labeled and unlabeled data in industrial processes, the semi-supervised strategy is introduced to learn fault-related features for better fault classification. To validate the performance of the proposed model, it is applied on the Tennessee–Eastman process and a real-world industrial hydrocracking process. The experimental results show that the proposed model has higher fault classification performance compared to other deep learning models. • A novel semi-supervised Inter-Relational Mahalanobis SAE (IRM-SAE) model is proposed to learn inter-relational distribution and fault-relevant dynamic features of process data. • An active time frame technique is developed to preprocess the process data to capture the dynamic features of the data. • To fully utilize both labeled and unlabeled data in industrial processes, the semi-supervised strategy is introduced to learn fault-related features for better fault classification. • The experimental results on two industrial processes show that the proposed model has higher fault classification performance compared to other deep learning models. [ABSTRACT FROM AUTHOR]

While Fisher discriminant analysis (FDA) has been widely used for classification, it highly relies on the label information of the training data, which means that its classification performance cannot be guaranteed if there are only a small number of labeled data samples available for use. For fault classification in industrial processes, unfortunately, there are always a much smaller number of faulty samples compared to the normal samples. In addition, even if the collected faulty samples are enough for modeling, it will still need expert experiences and prior process knowledge to label them into different types, which is time-consuming and costly. In this paper, a semi-supervised form of the FDA model is proposed and used for fault classification in industrial processes. The named semi-supervised FDA bridges the superior class separability of FDA and the unsupervised nature of principal component analysis (PCA). With the incorporation of additional unlabeled data samples for modeling, the fault classification performance has been greatly improved by the new model. Both of the semi-supervised modeling efficiency and the fault classification performance are evaluated through two case studies. [ABSTRACT FROM AUTHOR]

Abstract: This paper discusses how to use the warping information obtained after batch synchronization for process monitoring and fault classification. The warping information can be used for i) building unsupervised control charts or ii) fault classification when a rich faulty batches database is available. Data from realistic simulations of a fermentation process of the Saccharomyces cerevisiae cultivation are used to illustrate the proposal. [Copyright &y& Elsevier]