S171011 Abnormal Sounds Generation focused on Periodic Fluctuation for Rotary Machine Diagnosis
Kesaaki Minemura, Takashi Saeki, Shinya Yuda · The Proceedings of Mechanical Engineering Congress Japan · 2013
It is important to detect early abnormal sign to keep availability of industrial machines. Usually, we can't get abnormal operating sounds data because industrial equipment is normal operating state. We study the method of abnormal sounds synthesis focused on periodic fluctuation. Acoustic sensing data means convolution of sound source and diffusion. We developed the method of making abnormal sound after dividing sound source from diffusion. And we verify our method by simple experiment. In the result, we realized the improvement of abnormal sounds synthesis.