Numeric Simulation And Experimental Research Of Sound Source Identification With Beamforming Method

Acoustic signal carries abundant information about the working condition of the machine and it can be used in mechanical equipment monitoring and fault diagnosis for its convenience. Nowadays, Acoustic Fault Diagnosis has become one of the key research areas in mechanical equipment monitoring and fault diagnosis. To identify the sound sources and to find the causes of noise play an important role in acoustic fault diagnosis. Based on thorough analysis of various sound source identification methods and according to the project, beamforming method is selected to identify the sound sources. The summary of this paper is as follows:First, the background and the funded project of the research are discussed. After reviewing the development history of the sound source identification, the core content of this paper is established : the simulation and experimental research of sound source identification with beamforming method based on microphone array aiming to identify the sound sources of the working elevator.The theory and principle of beamforming are introduced in detail. The beamforming method based on spherical wave is applied in sound source identification in order to get more information. Three different algorithms including time domain method, frequency domain method, Auto-spectrum removal method are discussed and compared, with the conclusion that the time domain method can find the sound sources in the whole frequency band; the frequency domain method can "focus" on the sound source in a narrow frequency band and the Auto-spectrum removal method is more robust in low signal to noise ratio environment.