Data-driven analysis and optimization of low-frequency noise suppression in a multi-mechanism synergistic space-coiled structure

To address the engineering challenge of effectively suppressing low-frequency vibration and noise in shipborne equipment, a data-driven, multi-mechanism synergistic space-coiled acoustic metamaterial is proposed. By incorporating local resonant units and folding the acoustic propagation path, the proposed structure enables coordinated regulation of low-frequency sound waves, thereby enhancing low-frequency sound insulation performance. Sound transmission loss experiments are conducted to validate both the improved low-frequency sound insulation performance of the structure and the effectiveness of the proposed computational approach. Furthermore, the influence of key structural parameters on acoustic performance is systematically analyzed. On this basis, a Deep Multilayer Perceptron (Deep MLP) model, combined with a dung beetle optimizer (DBO) strategy, is introduced to optimize the critical structural parameters. The optimized acoustic metamaterial array is then implemented as an acoustic barrier within a cabin model for noise reduction evaluation. Results show that the optimized metamaterial reduces the average cabin sound pressure level by 4.01 dB over 150–300 Hz, resulting in a lower average SPL than the conventional acoustic barrier. This result indicates the acoustic attenuation effect of the proposed metamaterial under the considered simplified cabin conditions.