Unsupervised fault detection and diagnosis methods generally have the following shortcomings in their projection vectors: 1) they may not be specially designed to differentiate between normal and ...abnormal samples; 2) they may remain unchanged for different abnormal conditions; and 3) the key variables for the process anomalies may have not been effectively selected. In this study, a fault detection and diagnosis scheme sparse distribution dissimilarity analytics (SDDA) is proposed with a lasso penalty and distribution dissimilarity to solve these issues. The proposed method is formulated through a nonconvex optimization problem with a lasso penalty, which aims to maximize the distribution dissimilarity between different data sets. Then, the nonconvex optimization problem is recast to an iterative convex optimization problem using the minorization-maximization algorithm. After that, the constraint conditions are removed using the Karush-Kuhn-Tucker conditions for further simplification. Finally, the unconstraint optimization problem is solved through the proposed feasible gradient direction method. Based on the obtained sparse projection vectors, a fault detection model with both static deviation and dynamic fluctuation is developed. Since the statistics are designed using distribution dissimilarity, some abnormal conditions with small fault magnitudes can also be accurately detected. Besides, a reconstruction-based contribution (RBC) method is proposed for the statistics, and its diagnosability has been strictly demonstrated in theory. The detection and diagnosis performance of the proposed SDDA method is validated using a simulated process and a real industrial process. Experimental results illustrate the superiority of the proposed method to some commonly used methods.
Contemporary datasets can be immense and complex in nature. Thus, summarizing and extracting information frequently precedes any analysis. The summarizing techniques are many and varied and driven by ...underlying scientific questions of interest. One type of resulting datasets contains so-called histogram-valued observations. While such datasets are becoming more and more pervasive, methodologies to analyse them are still very inadequate. One area of interest falls under the rubric of cluster analysis. Unfortunately, to date, no dis/similarity or distance measures that are readily computable exist for multivariate histogram-valued data. To redress that problem, the present article introduces various dissimilarity measures for histogram data. In particular, extensions to the Gowda-Diday and Ichino-Yaguchi measures for interval data are introduced, along with extensions of some DeCarvalho measures. In addition, a cumulative distribution measure is developed for histograms. These new measures are illustrated for the Fisher iris data and applied to a U.S. temperature dataset.
We propose a new classification method for vector-valued images, based on: 1) a causal Markovian model, defined on the hierarchy of a multiscale region adjacency tree (MRAT), and 2) a set of ...nonparametric dissimilarity measures that express the data likelihoods. The image classification is treated as a hierarchical labeling of the MRAT, using a finite set of interpretation labels (e.g., land cover classes). This is accomplished via a noniterative estimation of the modes of posterior marginals (MPM), inspired from existing approaches for Bayesian inference on the quadtree. The paper describes the main principles of our method and illustrates classification results on a set of artificial and remote sensing images, together with qualitative and quantitative comparisons with a variety of pixel-based techniques that follow the Bayesian-Markovian framework either on hierarchical structures or the original image lattice.
The conventional multivariate statistical process control (MSPC) methods may not be sensitive to the detection of incipient changes since they in general quantify the distance between the new sample ...and the modeling samples without checking the changes of data distribution. In the present works, a method with dissimilarity analysis and quality-relevant subspace decomposition based on process monitoring method is developed to detect incipient abnormal behaviors that cannot be readily picked up by the conventional MSPC. First, the data are divided into quality-relevant subspace and the other subspace. Then dissimilarity analysis is performed to quantitatively evaluate the distribution difference between the normal condition and fault status for both subspaces. It can evaluate the incipient abnormal behaviors from the quality-relevant perspective to reveal the influences of incipient abnormality on quality. The paper demonstrates that the new method is more sensitive to the detection and isolation of incipient abnormal behaviors that are responsible for the distortion of the underlying covariance structure. Besides, it can tell whether the incipient fault can influence the quality index or not. Its feasibility and performance are illustrated with industrial process data.
PDF
