基于小波变换和神经网络的车内噪声信号重构

Interior noise signal reconstruction method based on wavelet transform and BP neural network

摘要:

为获取较高精度车内噪声主动控制(Active Noise Control, ANC)参考信号,提出了一种基于小波变换和BP神经网络的车内噪声信号重构方法。以在某轿车采集到的噪声信号为基础,用声学传递路径分析(TPA)方法确定影响车内噪声的关键点信号。鉴于噪声源信号对车内信号非线性关系的复杂性,建立BP神经网络的噪声重构模型,并利用小波分解来降低噪声信号的非平稳性。为对比重构效果,建立BP神经网络噪声重构模型。结果表明,本文提出算法的重构值与实测值之间的平均绝对误差比BP神经网络小,并且基于小波变换和BP网络重构模型的平均绝对误差均小于0.01。该方法能够对车内噪声信号进行准确、有效的重构。

Abstract:

To obtain high-precision active noise control (ANC) reference signal, a reconstruction method of interior noise signals that based on wavelet transform and BP neural network was proposed. Based on the noise signal sources collected in a vehicle, the key point signals affecting the interior noise were determined using the acoustic transfer path analysis (TPA) method. In view of the complexity nonlinear relationship between the noise source signals and interior signals, a BP neural network reconstruction model was established. And then wavelet decomposition method was used to reduce the non-stationarity of signals. Comparing the reconstruction effect, a BP neural network was established at the same time. The results show that the average absolute error between the proposed method reconstruction values and the measured values is smaller than that of the BP neural network. And the average absolute error of BP network reconstruction model based on wavelet transform is less than 0.01. This method can be used to reconstruct the noise signals on passenger ear-sides accurately and effectively.