Litcius/Paper detail

Multi-Sensor Fault Diagnosis for Misalignment and Unbalance Detection Using Machine Learning

Tauheed Mian, Anurag Choudhary, Shahab Fatima

20222022 IEEE International Conference on Power Electronics, Smart Grid, and Renewable Energy (PESGRE)19 citationsDOI

Abstract

Being the integrated part of any production or power generation system, rotating machines are the most evident for fault diagnosis. Nevertheless, with the rapid increment in competition among industries, there is a need for an intelligent multi-sensor based reliable diagnosis system for these machines. In the present study, the fault diagnosis using two sensors, namely a contact type vibration sensor and a non-contact type Infrared Thermal Imaging (IRT) camera, was utilized. For the analysis of misalignment and unbalance detection, the study is done for the optimal location of the vibration transducers and tested using three prevalent classification algorithms, viz. Naïve Bayes (NV), k-Nearest Neighbor (k-NN), and Support Vector Machine (SVM) for fault classification and performance evaluation. However, the results based on IRT are found to be 100% for the diagnosis of considered faults using SVM and leads to an effective way of fault diagnosis in rotating machines.

Topics & Concepts

Computer scienceFault detection and isolationFault (geology)Artificial intelligenceMachine learningPattern recognition (psychology)ActuatorGeologySeismologyFault Detection and Control SystemsMachine Fault Diagnosis TechniquesElectrical Fault Detection and Protection
Multi-Sensor Fault Diagnosis for Misalignment and Unbalance Detection Using Machine Learning | Litcius