Bearing Fault Analysis Utilizing Fuzzy Logic Methodology for Enhanced Diagnostic Accuracy

This research aims to design a tool that can be used to detect damage or malfunctions in induction motors, especially in the bearing part which is the main driving component. Using the MPU 6050 accelerometer sensor and sound sensor module with the Arduino Nano microcontroller and the HC 05 Bluetooth module as a medium for sending and acquiring signal data. The signal data obtained in the form of a datalog with the extension .txt is then processed further with Matlab software to find out information on the characteristics of sound signals and vibration signals generated by induction motor bearings. Using a cut-off signal filter low pass filter for voice signal filter processing and fast Fourier transform (FFT) to convert the time domain signal into a signal frequency domain to determine the frequency characteristics arising from the signal. From the sound and vibration signal input data obtained, the Fuzzy logic method is used to determine the bearing condition output. The developed system is capable of detecting bearings in three conditions, namely good, damaged, and alert.