基于实验SEA方法的车内噪声预测分析
Predictive Analysis of Automobile Interior Noise Based on Experimental SEA Method

针对车内中高频噪声预测分析准确性低的缺陷，提出了一种有效的基于实验分析的有统计能量分析(statistical energy analysis,简称SEA)方法。首先,利用点导纳平均法得到速度与力的实部，确定模态密度；其次,采用衰减法得到脉冲响应，经过Hilbert变换推论出内损耗因子；最后,利用异点导纳平均法，测得带宽内的子系统的平均能量，计算出耦合阻尼损耗因子。依托某车中高频降噪分析测得某工况下的输入声载荷，将实验测得的SEA基本参数输入到模型中，得到驾驶员右耳处的噪声水平，其仿真值与测试值在中高频时吻合较好，证明了轿车模型的有效性。结果表明，基于实验方法测得SEA3个基本参数可以综合提高模型预测车内噪声的准确性，具有很高的工程应用价值。
A method based on the experimental statistical energy analysis(SEA) is introduced to amend the defects of predicting the interior noise at medium-high frequency. First, modal density is derived from the input mobility, and the damping loss factor is deduced based on the decay rate method and Hilbert transform. Then, the coupling damping loss factor is calculated from the band average input power which is measured based on the transfer function from each drive point to all receiver points. Finally, the SEA parameters measured in the experiments are imported into the model for the simulation of the noise around the driver′s right ear. The comparison shows that the simulation results are in coincident with the measurements, which proves the validity of model based on experimental SEA. It is of great value in engineering application on improving the accuracy of interior noise prediction on the base of experimental SEA method.