A broadband low-frequency muffler based on neural network method and Helmholtz resonator with helical neck

In this study, a design method of broadband low-frequency muffler based on NN method and Helmholtz resonators with helical necks is proposed, which ensures almost perfect ventilation and broadband sound insulation in the low- and medium-frequency range. First, the sound insulation performance of the muffler is improved by adding a helical structure at the port. The slit Helmholtz resonator with a helical structure is able to reduce the cross-sectional area of the resonator port and increase the contact area of thermal viscosity loss, thus reducing the peak frequency of sound insulation and increasing the sound energy loss. Using finite element analysis software, the structural transmission loss is simulated. At the same time, the neural network method is used to predict the combined structure, then, the combined structure can be designed quickly and reasonably to achieve the sound insulation of medium- and low-frequency broadband. Experimental and simulation results show that in the target range of 430 Hz-2220 Hz, the overall TL exceeds 30 dB and the maximum exceeds 60 dB. In addition, when a helical structure is added, the overall pipeline remains unchanged; and when a muffler is added on the basis, the perfect ventilation effect can be achieved. The proposed structural model and the idea of rapid design lay a great foundation for the design and optimization of broadband muffler.