No abstract

Aiming at the problem that there is a strong eardrum pressure in the passenger car during the closing process, two analysis and prediction methods, fast formula prediction and CFD simulation based on accurate models, are proposed. The regression model of ear pressure comfort was established by DOE method and multiple linear regression; The simulation software star-CCM+ is applied to simulate and analyze the dynamic characteristics of the flow field in the cockpit during the closing process by using the overlapping grid technology, and the pressure change curve near the ear is obtained. Finally, the CFD numerical simulation model is established by comparing and analyzing the regression prediction analysis results and the real vehicle test data. The results show that the effects of closing speed, effective opening area of pressure relief valve and air tightness of the whole vehicle on the pressure of passengers’ eardrums decrease in turn, and the prediction error of multiple linear regression equation is 17 %; The analysis error of the internal flow field dynamic characteristic model based on refined modeling is 8 %. This study provides a theoretical basis for solving the problem of rapid prediction of eardrum pressure and optimization of engineering structure.

No abstract

To accurately identify the abnormal door-closing noise, we propose a method to recognize the time-frequency image of door closing sound based on a multi-scale feature fusion network model. The door-closing sound signal is transformed into a time-frequency image through wavelet analysis, and a classification model based on multi-scale feature fusion is designed. The model introduces multi-scale filters and dilated convolution and adds two improved inception modules to keep the model lightweight. At the same time, richer spatial features can be obtained. The features of different scales are spliced and input to the fully connected layer, and a dropout layer is added to the fully connected layer to suppress overfitting. By comparing the loss and accuracy rate, adjusting different hyperparameters, the optimal model is obtained. The experimental results show that the multi-scale feature fusion network model has a higher accuracy rate than the transfer learning model. Test accuracy rate is 86% and can effectively recognize abnormal door-closing noise. It provides a feasible theoretical basis for the direction of abnormal door noise recognition.