MATLAB-based fault diagnosis of industrial rotor-bearing systems

Abstract

Continuous monitoring of critical rotor-bearing systems is crucial in order to prevent machine downtime, which would otherwise lower the overall output and quality. Complex modern machinery demands an upgraded intelligent fault diagnosis method that leaves minimal room for human error. This paper presents a MATLAB-based condition monitoring and fault diagnosis method for rotating machines used in sugar factories. The vibration responses are acquired through the use of data acquisition and the fast Fourier transform (FFT) analyser on real industrial machines. These signals are supplied as the input to a specially developed MATLAB program for processing in order to detect the fault and help to suggest remedies. The simple and user-friendly approach saves time and increases the effectiveness of condition monitoring in the reduction of downtime and the avoidance of catastrophic failure in industrial machines.