音響診断による設備監視効率化の検討
概要
In order to investigate a method for efficiently monitoring multiple machines in a building with an acoustic sensor, an experiment was performed to acquire acoustic data of three rotating machines running simultaneously in a state in which all three machines are normal or one of three is abnormal, with two microphones placed at different positions. The abnormality can be detected from the acoustic signal level, and it can also be specified that the bearing abnormality has occurred by frequency analysis. A method to identify the machine in which the abnormality occurred from the ratio of the increase amount of the signal level of the two microphones was proposed, which showed the ability of identifying the abnormal machine correctly under 5 conditions out of 6 conditions
