Active noise control technique using the neural network for ventilation openings of building

在住宅中保持安静对于舒适的生活环境很重要。通风口可能成为住宅外墙隔声性能的瓶颈。在本研究中，提出了一种新的神经网络算法，用于通过通风口对噪声进行主动控制。通过数值模拟检验，比较了传统的LMS算法和NN算法的ANC效果。结果表明，反复研究噪声模式的神经网络算法比LMS算法具有更高的性能。

Keeping silence in residences is important for comfortable living environment. Ventilation openings might be a bottleneck of sound insulation capability of exterior walls in the residence. In this study, new algorithm applying the neural networks was proposed for active control of the noise through the ventilation openings. By examinations using numerical simulations, ANC effects by the NN algorithm was compared to effects by LMS algorithm that was an orthodox method. The results demonstrated that the NN algorithm that had studied noise patterns repeatedly had higher performance than the LMS algorithm.