Random forest-based nonlinear improved feature extraction and selection for fault classification

Abstract

In this paper, Interval Gaussian Process Regression (IGPR)-based Random Forest (RF) proposed for fault detection and diagnosis (FDD) due to its effectiveness in handling uncertain industrial process data, which are often with high nonlinearities and strong correlations. This technique aims to extract the features from raw data using IGPR technique. Then, the interval mean vector and the interval variance matrix obtained from IGPR technique are used as inputs to the Random Forest (RF) classifier. The results show the effectiveness of the features and the classifiers in detection of faults of Wind Energy Conversion (WEC) Systems.