R. Ahila, M. Sudhakaran, and K. Manimala
Wavelet transform, Parseval’s theorem, power quality, support vector machine, feature selection
Recognition of the presence of any disturbance and classifying any existing disturbance into a particular type is the first step in combating the power quality (PQ) problem. In many pattern recognition applications, high-dimensional feature vectors impose a high computational cost as well as the risk of "overfitting". Feature extraction and feature selection are two different approaches for the reduction of dimensionality. Wavelet transform (WT) has been used to extract some useful features of the power system disturbance signal. Feature selection addresses the dimensionality reduction problem by determining a subset of available features which is most essential for classification. This paper presents a novel hybrid filter wrapper type feature selection method named filtered and wrappered sequential forward search (FW_SFS) in the context of support vector machines (SVM). In comparison with conventional wrapper methods that use the SFS strategy, FW_SFS has two important properties to reduce the time of computation. First, it dynamically maintains a subset of samples for the training of SVM. Because not all the available samples participate in the training process, the computational cost to obtain a single SVM classifier is decreased. Secondly, a new criterion, which takes into consideration both the discriminant ability of individual features and the correlation between them, is proposed to effectively filter out non-essential features. As a result, the total number of training is significantly reduced and the overfitting problem is alleviated. Results of simulation and analysis demonstrate that the proposed method can achieve higher correct identification rate, better convergence property and less training time compared with the method that use the full feature set.
Important Links:
Go Back
