State Estimation for Sound and Vibration Emitted from aMachine Based on Sound Measurement Under Existenceof Background Noise

Many studies on the diagnosis for machines have become important recently because of increased use of various complex industrial systems. The correlation information between sound and vibration is very important for machine diagnosis. Usually, vibrationpickups are attached directly to the machine in order to measure vibration data. However, in some cases, the sensors can not be attacheddirectly on highly precise devices. In this study, a method to estimate the fluctuation of sound and vibration is proposed based on themeasurement data of sound emitted from the machine under existence of background noise. The effectiveness of the proposed theoryis experimentally confirmed by applying it to the observed data emitted from a rotational machine driven by an electric motor.