Numerical Optimization of Single-chamber Mufflers Using Neural Networks and Genetic Algorithm


Abstract: To simplify the optimization process, a simplified mathematical model of a muffler is constructed using a neural network and a series of input design data (muffler dimensions) and output data (theoretical sound transmission loss) that are obtained by utilizing a theoretical mathematical model (TMM). To assess an optimal muffler, a neural network model (NNM) is used as the objective function in conjunction with a genetic algorithm (GA). Before the GA operation can be carried out, however, the accuracy of the TMM must be checked and be in accord with the experimental data. Additionally, the NNM must also be in agreement with the TMM. Also discussed are the numerical cases of sound elimination relative to the various parameter sets and pure tones (500, 1000, and 2000 Hz). The results reveal that the maximum value of the sound transmission loss (STL) can be accurately obtained at the desired frequencies. Consequently, the algorithm proposed in this study can provide an efficient way to develop optimal silencers for the requisite industries.

Keywords: Four-pole transfer matrix, Polynomial neural network model, Optimization, Genetic algorithm

Full Text: PDF