Radial Basis Function Neural Network for prediction of medium-frequency sound absorption coefficient of composite structure open-cell aluminum foam

Abstract In order to better understand the acoustic performance of composite structure open-cell aluminum foam at medium-frequency bands. In this paper, we propose a method that used the Radial Basis Function Neural Network to predict the sound absorption coefficient of the given structure at medium-frequency bands. The condition of the method is that the acoustic performance of composite structure open-cell aluminum foam is determined by some structural parameters which can be easily measured. Therefore, in experiment section, 16 composite structure open-cell aluminum foam are designed with three types of NO.1 open-cell aluminum foam and 3 types of NO.2 open-cell aluminum foam firstly. Secondly, five structural parameters, porosity, pore size, thickness, cavity and density of each layer were measured as the input of Radial Basis Function Neural Network. Thirdly, the sound absorption coefficient of each composite structure at 500 Hz, 800 Hz, 1000 Hz, 1250 Hz, 1600 Hz are measured with AWA8551 impedance tube as the output of Radial Basis Function Neural Network. Finally, the prediction results and the maximum relative error is 5.7% of five samples from no.12 to no.16 show that the method is effective.