IN SITU ESTIMATION OF THE SURFACE ACOUSTIC IMPEDANCE IN REALISTIC INTERIORS BY AN ACOUSTICAL INVERSE APPROACH

Methods for in situ measurement of acoustic impedance and absorption coefficients have gained importance in the last years as a result of the advances in the computational techniques to analyze and predict sound fields, which have lead to a closer dependency between the quality of acoustic properties assessment and the accuracy of the predictions. Among the existent methods for in situ measurement, those based on inverse boundary formulations offer the advantage of being able to deal with surfaces of arbitrary shapes. However, they suffer from high sensitiveness to noise due to the solution of a rank deficient linear model. The inverse approach presented in this report avoids such difficulty by transforming the rank deficient inverse model into an iterative optimization procedure in virtue of the a priori knowledge of the geometry of the surfaces in a room. Thus, the input data consists of the geometric model of the room, the strength of a harmonic sound source, and the sound pressure measured at a number of arbitrary points in the interior field. With this information, the iterative algorithm estimates the normal acoustic impedance of not only one but all the surfaces even in the presence of noise, as demonstrated by numerical simulations with different signal to noise ratios. In practice, the measurement process is accomplished with a single microphone moving freely in space and tracked by stereoscopic video cameras. In this paper are presented the results of an attempt to estimate the acoustic impedance of the prearranged surfaces set on the walls of an experimental acoustic room.