轿车车窗上升过程声品质预测模型研究  

A Sound Quality Prediction Model for the Car Window Raising Process

作　　者：李文平[1] 何川林 张帅 田亚坤 LI Wenping;HE Chuanlin;ZHANG Shuai;TIAN Yakun(School of Vehicles and Energy,Yanshan University,Qinhuangdao 066004,Hebei,China)

机构地区：[1]燕山大学车辆与能源学院,河北秦皇岛066004

出　　处：《汽车工程学报》2023年第3期373-379,共7页Chinese Journal of Automotive Engineering

摘　　要：针对声品质主观评价试验成本高的问题,提出一种基于车窗上升过程稳态噪声和瞬态噪声的声品质预测模型。采集24款轿车左前门车窗上升过程的噪声信号,在时域上划分为具有瞬态噪声特征的启动阶段、停止阶段和具有稳态噪声特征的平稳运行阶段,分别计算出3个阶段的客观评价参数。通过主观评价试验得到每款样本车的主观评价结果。运用BP神经网络建立了车窗上升过程的声品质预测模型,预测结果表明,此方法提取的客观评价参数能反映左前门车窗上升过程的声品质特征,BP神经网络建立的声品质预测模型具有较高准确度和泛化能力,能在一定程度上代替评测员对车窗上升过程声品质进行评价。To address the issue of high cost of subjective evaluation experiment of sound quality,the paper proposes a prediction model for sound quality based on both steady and transient noise during the windowraising process.Noise signals were collected from the left front door window of 24 cars while raising the window.These signals were divided into the start and stop stages which exhibit transient noise characteristics,and the steady operation stage with steady noise characteristics.The objective evaluation parameters of the three stages were calculated respectively.The subjective evaluation results for each sample car were obtained through corresponding experiments.A sound quality prediction model was established using a BP neural network.The prediction results show that the objective evaluation parameters extracted using this method effectively represent the sound quality characteristics during the raising process of the left front door window.The sound quality prediction model based on the BP neural network has high accuracy and generalization capabilities,allowing it to replace the human evaluators to a certain extent in assessing the sound quality during the window raising process.

关 键 词：声品质 主观评价 预测模型 稳态噪声 瞬态噪声 

分 类 号：U467.493[机械工程—车辆工程]