Effectiveness of Entropy in Equipment Diagnosis for Acoustic Data

Abstract

A method for recording and analyzing sounds generated by target equipment with a smartphone for equipment diagnosis is under investigation. In this report, we focused on the amplitude change caused by the sound generated by a bearing flaw, and found the possibility of detecting abnormalities using entropy, mutual information content, and Kalbach-Leibler divergence. However, the amplitude is affected by the distance between the smartphone and the equipment, and we are currently considering how to deal with this issue. The goal is to make it possible for anyone to easily perform inspections with a single small terminal.