Evaluation Method of Instantaneous Acoustic Quality of Automobile Latch Based on BP Neural Network

Aiming at improving the acoustic quality of the latch mechanism of automotive side door, the contact point of the car side door slam with the loudest sound is obtained. In view of the difficulty of obtaining sufficient experimental instantaneous acoustic samples during the door slam process of latch acoustic testing, a method combined with FEM and BEM using ABAQUS and LMS software is utilized skillfully to obtain acoustic samples. The credibility of the simulation samples is verified within an error range smaller than 12%. Finally, a subjective-objective acoustic quality evaluation method and algorithm based on BP neural network is realized in MATLAB. The prediction results in BP neural network are tallied with the real results well. This study provides a reference for the objective evaluation of latch acoustic quality during process of door slam.