Evolving Fuzzy Classifier based on Clustering Algorithm and Drift Detection for Fault Diagnosis Applications

Abstract

Nowadays, in several areas, efficient fault diagnosis methods for complex machinery and equipments are required. Several fault diagnosis methods based on different theories and approaches have been proposed in the literature. In general, these methods use mathematical/statistical models, accumulated experience, or even process historical data to perform fault diagnosis. Although methods based on models or experience have shown to be effective, they have the disadvantage of requiring previous knowledge of the dynamic system in question. On the contrary, methods based on process historical data do not require a prior knowledge, they are based solely on data obtained directly from the dynamic system. The application of so-called “Evolving Intelligent Systems” to accomplish fault diagnosis from process data have been shown a promising approach. This paper proposes an evolving fuzzy classifier based on a new approach that combines a recursive clustering algorithm and a drift detection method and its application on dynamic systems fault diagnosis. The novel approach provides greater robustness to outliers and noise present in data from process sensors. The classifier is evaluated in fault diagnosis of an interacting tank system and the results are promising.