Search

Independent component analysis (ICA) is a feature extraction technique for blind source separation. Enhanced independent component analysis (EICA), which has enhanced generalization performance, operates in a reduced principal component analysis (PCA) space. SVM is a powerful supervised learning algorithm, which is rooted in statistical learning theory. SVM has demonstrated high generalization capabilities in many pattern recognition problems. In this paper, we integrate EICA with SVM and apply this new method to the leak detection problem in oil pipelines. In features extraction, EICA produces EICA features of the original pressure images. In classification, SVM classified the EICA features as leak or non-leak. The test results based on real data indicate that the method can detect many leak faults from a pressure curve, and reduce the ratio of false and missing alarm than conventional methods. [ABSTRACT FROM AUTHOR]

Fusion of the states of a nonlinear dynamic multiscale system (DMS) on the basis of available noisy measurements is one of the well-known key problems in modern control theory. To the best of our knowledge, all of the previous work focused attention on linear DMS. However, nonlinear DMS has never been investigated. In this paper, modeling and fusion estimation of dynamic multiscale system subject to nonlinear measurement equation is proposed. Haar wavelet is used to link the scales. Monte Carlo simulation results demonstrate that the proposed algorithm is effective and powerful in this kind of nonlinear dynamic multiscale system estimation problem. [ABSTRACT FROM AUTHOR]

This paper proposes an improved fuzzy neural system for very-short-term load forecasting problem, based on chaotic dynamics reconstruction techniques. The Grassberger-Procaccia algorithm and the Least Squares Regression method were applied to obtain the accurate value for the correlation dimension, which is used as an estimation of the model order. Based on that model order, an appropriately structured Fuzzy Neural System (FNS) for load forecasting was designed. Satisfactory experimental results were obtained in 15 minutes ahead electrical load forecasting from the electric utility in Shandong Heze area. And the same experiments using conventional Artificial Neural Network are also performed as a comparison with the proposed approach. [ABSTRACT FROM AUTHOR]