A sound insulation prediction model for floor structures in wooden buildings using neural networks approach

Reliable prediction tools are yet to be developed for estimating the accurate acoustic performance of lightweight structures, which are vital especially in the design process. This paper presents a sound insulation prediction model based on artificial Neural Networks (NN) to estimate acoustic performance for airborne and impact sound insulation of floor structures. At an initial stage, the prediction model was developed and tested for a small amount of data, specifically 67 laboratory measurement curves in one third octave bands. The results indicate that the model can predict the weighted airborne reduction index R_w for various floors with a maximum error of 1 dB. The accuracy decreases with errors up to 9 dB for the weighted index for impact sound L_n,w, in cases of complex floor configurations due to large error deviations in high frequency bands between the real and estimated values. The model also shows a very good accuracy in predicting the airborne and impact sound insulation curves in the low frequencies, which are of higher interest usually in building acoustics.