Data-driven approach to machine condition prognosis using least square regression tree

Abstract

Machine fault prognosis techniques have been profoundly considered in the recent time due to their substantial profit for reducing unexpected faults or unscheduled maintenance. With those techniques, the working conditions of components, the trending of fault propagation, and the time-to-failure are precisely forecasted before they reach the failure thresholds. In this work, we propose an approach of Least Square Regression Tree (LSRT), which is an extension of the Classification and Regression Tree (CART), in association with one-step ahead prediction of time-series forecasting techniques to predict the future condition of machine. In this technique, the number of available observations is firstly determined by using Cao's method and LSRT is employed as prediction model in the next step. The proposed approach is evaluated by real data of low methane compressor. Furthermore, a comparison of the predicted results obtained from CART and LSRT are carried out to prove the accuracy. The predicted results show that LSRT offers a potential for machine condition prognosis.

Statistical Aspects in Neural Network for the Purpose of Prognostics

“…It includes NN (Chakraborty, Mehrotra, Mohan & Ranka, 1992;Yao, 1999), Gaussian process regression (Seeger, 2004;Mohanty, Teale, Chattopadhyay, Peralta & Willhauck, 2007), relevance vector machine (Tipping, 2001), least square regression (Tran & Yang, 2009), etc. Among these algorithms, NN is a representative data-driven method, in which a network model learns a way to produce a desired output such as future damage level reacting to given inputs such as previous damage level and usage conditions instead of physical model.…”

Prognosis of a Wind Turbine Gearbox Bearing Using Supervised Machine Learning

“…The Logistic regression model was employed in [15] for the estimation of RUL of CNC machine. More regression-based prognostic models can be found in the literature [16,17,18,19,20].…”

Evaluation of Flashover Voltage Levels of Contaminated Hydrophobic Polymer Insulators Using Regression Trees, Neural Networks, and Adaptive Neuro-Fuzzy

“…Correllation analysis will be employed to determine such relation [22]. Consequrntly, this research work investigates three modern artifitial intelligent regression models: regression trees, artifitial neural netwoks, and adaptive neuro-fuzzy [23][24][25].…”

“…In [25], regression tree was used as a base for data-driven to machine condition prognoses. In [26], it was used to predict the wind speed with local models or algorithms being discussed for the short term operating wind conditions.…”

“…The superiority and effectiveness of the regression trees are achieved in [25][26][27][28][29][30][31][32][33][34][35]. Furthermore, it is clear from the literature survey that the application of regression trees algorithm of flashover voltage prediction has not been discussed.…”