Hirokazu ONO

Research on Analysis Methods Used For the Noise Propagation Model.

Yasuyuki MIYAKI

The noise is one of the important environmental problems.
In this research, noise propagation prediction is performed using the Cellular Automata method (the following CA method). The last year, the development of two-dimension prediction was performed. As a result, prediction in easy two-dimension problem was completed. However, various phenomena occur in the noise propagation. Therefore each examination was required enough against each phenomenon. And, the calculated result of the CA method wasn't compared with the result of other analysis method.
Then, the prediction is performed using the finite difference method in this year. And, the calculated results of finite difference method and the CA method are compared. These compare is calculated in one- and two-dimension propagation. The one-dimensional problem is calculated for the acoustic pipe model. The two-dimensional problems are calculated for the divergence decrease, the diffraction attenuation, the interference and the Doppler effect. The calculated results of the CA method are the same as those of the finite difference method.
Next, we examine the conditions which an exact sound pressure value is predicted by the CA method. The condition is examined by the relation between the frequency and the width of cell. Moreover, the calculated time of the CA method and the finite difference method are examined.
As a result, we observe that noise propagation prediction of the CA method is reliable.