Adaptive Active Vehicle Interior Noise Control Algorithm Based on Nonlinear Signal Reconstruction

In this study, to reduce secondary sound source pollution in the reference signal of active noise control (ANC), a novel ANC algorithm, based on signal reconstruction, is proposed for vehicle interior noise. This algorithm combines the processes of ear-sides noise reconstruction and ANC. First, to reduce non-stationarity and nonlinearity, multi-source noise signals outside the vehicle are decomposed into a finite number of intrinsic mode function (IMF) components by empirical mode decomposition (EMD). Second, the IMFs are reconstructed by the energy-extreme division method into three components: high-frequency, intermediate-frequency and low-frequency. The radial basis function neural network (RBFNN) parameters are adjusted by the proportions of the components. Model training is performed to obtain the high-precision EMD–RBFNN reconstruction model (EMD–NNRM). The reconstructed noise signal is used as the reference signal of the variable step-size least mean square (VSS-LMS) algorithm, to control the passenger ear-sides noise. The effectiveness of the EMD–NNRM is validated using four noise signals from the outside of a vehicle. The interior noise of a high-speed vehicle is processed by the proposed algorithm and the traditional VSS-LMS algorithm for comparison. The reconstruction results show that the mean absolute error is improved by 77.64% compared with the back propagation neural network reconstruction model. Reconstructed passenger ear-sides noise can be utilized for ANC. The active control results suggest that the proposed algorithm can not only effectively suppress the interior noise but can also avoid pollution from secondary sound sources.