Natural gas boiler monitoring system based on vibration analysis using wavelet decomposition

Abstract

In this paper, a wavelet application which aims to monitor a domestic natural gas boiler as a mechanical system is presented. In this predictive maintenance application, discrete-time wavelet decomposition is basically used to analyze the recorded vibration signal of the mechanical system. A low cost microphone is utilized to capture the vibrations. Then a Simulink model of the application is built. Thus, real-time observations become possible. This approach contributes developing intelligent gas boilers that observe their own situation.