Abstract
The rapid growth of science and technology in modern civilisation has led to an increase in the size, complexity, and automation of machinery and equipment. Two of the most important aspects of modern industrial production are problem identification and machinery condition monitoring. Early problem detection is made possible by effective condition monitoring, which is crucial when taking into account variables like production efficiency, operational dependability, maintenance costs, and downtime. Research on the identification of problems and the health monitoring of machinery has practical implications. For the purposes of equipment monitoring and fault diagnostics, information on the temperature, vibration, noise level, and lubrication state of the equipment is recorded. After that, the information is utilised to identify the issue’s primary source and put remedial measures in place. A condition monitoring system’s core elements are fault prediction, feature extraction, and problem diagnostics. Feature extraction and fault diagnostics are essential for normal detection, problem localisation, and failure severity prediction. This paper includes fault diagnosis and applications of computational intelligence in condition monitoring and fault detection also this paper presents a method for equipment status monitoring using Machine Learning (ML) techniques. Popular machine learning (ML) classification methods like Random Forest (RF), Random Tree (RT), Naive Bayes (NB), XG Boost (XGB), and Logistic Regression (LR) are used for assembling. The pressing need to increase machine reliability and reduce the possibility of production losses due to machine breakdowns is the reason behind the growing emphasis on machine condition monitoring.
Keywords: Failure Prediction, Machine Learning, Naive Bayes, Production, Random Forest, Random Tree, Unexpected Downtime.