Machine fault diagnosis and condition prognosis using classification and regression trees and neuro-fuzzy inference systems

This paper presents an approach to machine fault diagnosis and condition prognosis based on classification and regression tress (CART) and neuro-fuzzy inference systems (ANFIS). In case of diagnosis, CART, which is one of the decision tree methods, is used as a feature selection tool to select pertinent features from data set and ANFIS is used as a classifier. The crisp rules obtained from CART are then converted to fuzzy if-then rules that are employed to identify the structure of ANFIS classifier. The hybrid of back-propagation and least squares algorithm are utilized to tune the parameters of the membership functions. The data sets obtained from vibration signals and current signals of the induction motors are used to evaluate the proposed algorithm. In case of prognosis, both of these models in association with direct prediction strategy for long-term prediction of time series techniques are utilized to forecast the future values of machine’s operating condition. In this case, the number of available observations and the number of predicted steps are initially determined by using false nearest neighbor method and auto mutual information technique, respectively. These values are subsequently utilized as inputs for prediction models. The performance of the proposed prognosis system is then evaluated by using real trending data of a low methane compressor. A comparative study of the predicted results obtained from CART and ANFIS models is also carried out to appraise the prediction capability of these models. The results of the proposed methods in the two cases indicate that CART and ANFIS offers a potential for machine fault diagnosis as well as for condition prognosis.