No abstract

To reduce the intake noise of automobile engines, an active control system model of engine intake noise is established with the Filtered-x least mean square (FxLMS) algorithm. The offline identification method is adopted to identify the secondary path. The engine speed signal is used to construct the reference signal of a sound source to avoid interference of a secondary sound source to the reference signal. A variable-step algorithm is proposed, in which parameters are added to the normalized algorithm instead of the sinusoidal variable-step algorithm to adjust the amplitude range. This algorithm not only has the advantages of fast convergence speed and small steady-state error but also adapts to the characteristics of a time-varying reference signal and easy selection of parameters. In this paper, the noise of automobile engines under the New European Driving Cycle (NEDC) is studied and the proposed algorithm has faster convergence speed compared with the normalization algorithm, better adaptability to the change of the reference signal, and better stability compared with the sinusoidal variable step-size algorithm. The results show that the algorithm proposed can effectively reduce the intake noise of the engine at each speed and the noise reduction effect can reach 23.11 dB at a certain frequency. Meanwhile, the stability of the system is improved.

The speech enhancement effect of traditional deep learning algorithms is not ideal under low signal-to-noise ratios (SNR). Skip connections-deep neural network (Skip-DNN) improves the traditional deep neural network (DNN) by adding skip connections between each layer of the neural network to solve the degradation problem of DNN. In this paper, the Multiresolution Cochleagram (MRCG) features in the gammachirp transform domain are denoised to obtain the improved MRCG (I-MRCG). The noise reduction method adopts the Minimum Mean-Square Error Short-Time Spectral Amplitude Estimator (MMSE-STSA) and takes I-MRCG as the input feature and Skip-DNN as the training network to improve the speech enhancement effect of the model. This paper also proposes an improved source-to-distortion ratio (SDR) loss function. When the loss function uses the improved SDR, it will improve the performance of Skip-DNN speech enhancement model. The experiments in this paper are performed on the Edinburgh dataset. When using I-MRCG as the input feature of Skip-DNN, the average perceptual evaluation of speech quality (PESQ) is 2.9137, and the average short-time objective intelligibility (STOI) is 0.8515. Compared with MRCG as Skip-DNN input features, the improvements are 0.91% and 0.71%, respectively. When the improved SDR is used as the loss function of the speech model, the average PESQ is 2.9699 and the average STOI is 0.8547. Compared with other loss functions, the improved SDR has a better enhancement effect when used as the loss function of the speech enhancement model.