The Description of Sound Sources by a Set of Equivalent Sources Using Impulse Response Functions and the Least Mean Square (LMS) Algorithm (vor Ort)
* Presenting author
Abstract:
Often we are interested to describe the sound radiation from complex sources with a limited set of equivalent sources. Typical examples would be the sound radiation from vibrating structures or from the mouth of a human speaker. The main challenge is to find source distributions and source strengths in an appropriate way that directivity and strength in the farfield will be approximated with acceptable accuracy. To achieve this a time domain approach is suggested. Impulse response functions are used between equivalent sources and observation positions of the field to be reconstructed. The Least Mean Square (LMS) algorithm is applied to determine the source signals of the equivalent sources so that the error between the reconstructed field and the observed field in the observation is minimized. As the approach does not require matrix inversion it is very robust with respect to measurement errors (e.g. due to measurement noise). The approach is demonstrated for sound radiation from a vibrating plate based on monopole sources situated in front of the plate and for the case of a loudspeaker element with a set of monopoles simulating the loudspeaker. Results, advantages and disadvantages of the method are discussed in the presentation.
