Because of the position of the motor and the excitation of the suspension system, a wheel-driven electric bus produces low-frequency noise, which is difficult to resolve through traditional sound absorption and noise reduction technology. Through an interior noise test of a wheel-driven electric bus, we found that the interior low-frequency noise had a considerable influence on the driver. In order to solve this problem, a locally resonant phononic crystal was used to meet the requirements of vibration and noise reduction for the wheel-driven electric bus. The intrinsic relationship between the band gap distribution of the locally resonant phononic crystal and the topology was established by training a neural network, so as to achieve the desired effect of the bandgap model on the basis of the input bandgap range. Upon an increase in the number of models, the prediction model error decreased gradually. This method could quickly obtain the structural parameters of the locally resonant phononic crystal with the expected band gap, which made it convenient to apply locally resonant phononic crystals to the vibration and noise reduction of wheel-driven electric buses and in other fields.

由于电机的位置和悬架系统的激励，轮式电动客车产生低频噪声，这是传统吸声降噪技术难以解决的。通过对一辆轮式电动客车的车内噪声测试，我们发现车内低频噪声对驾驶员的影响相当大。为了解决这个问题，采用局部谐振声子晶体来满足轮驱动电动客车的减振降噪要求。通过训练神经网络建立局部共振声子晶体带隙分布与拓扑的内在关系，从而在输入带隙范围的基础上达到带隙模型的预期效果。随着模型数量的增加，预测模型误差逐渐减小。该方法可以快速获得具有预期带隙的局域谐振声子晶体的结构参数，便于将局域谐振声子晶体应用于电动公交车的减振降噪等领域。