车内低频噪声预测模型的改进

Model Improvement for Car Interior Low-frequency Noise Prediction

基于模态阻尼识别和载荷识别原理,提出了改进的车内低频噪声预测模型,结合实测加速度响应频谱,采用遗传算法修正模态阻尼比和激励力参数。将该方法应用于某轿车车内低频噪声预测中,建立包含座椅和驾驶员的轿车有限元-边界元声固耦合改进模型,对20~200Hz频带范围内车内噪声进行预测计算。同时通过试验测试实车怠速、30,40和50 km/h匀速工况下动力总成悬置点、车身悬挂点加速度信号和车内声压响应。结果表明：改进后的模型预测值与试验值吻合良好,预测精度优于传统模型。

Based on the principles of modal damping identification and load identification, an improvedmodel for car interior low-frequency noise prediction is put forward, with modal damping ratio and excitation forceparameters revised by using genetic algorithm and based on the frequency spectra of acceleration response measuredin real vehicle tests. The scheme is applied to the interior low-frequency noise prediction of a car. An improvedFEM-BEM acoustic-solid coupling model including seats and driver is established and the predicted interior noises inthe frequency range of 20-200Hz are calculated. Meanwhile corresponding tests are conducted to measure the accelerationsignals and interior sound pressure responses at engine mounting points and car body suspended points. Theresults show that the prediction values with improved model are well agree with test data, with a prediction accuracysuperior to the traditional model.