基于结构加速度误差最小化的声学预测模型修正

针对FEA(有限元分析)声学预测方法,提出了基于结构加速度响应误差最小化的声学预测模型修正方法。以模态迭加法为基础,结合模态应变能模态阻尼识别和频域载荷识别原理,重构了FEA声学预测模型,通过遗传算法优化结构损耗因子,使结构加速度响应计算值与实验值误差最小化,从而识别系统模态阻尼比及激励力,实现对声学预测模型参数的修正。将该方法与传统计算方法预测结果对比,结果表明:该方法可以精确识别系统模态阻尼和激励力,提高模型预测精度。

Modified model for acoustic prediction based on acceleration error minimization

A model modifying method for FEA (finite element analysis) acoustic prediction based on the minimization of structural acceleration response error is presented. FEA acoustic prediction model is reconstructed in combination with modal damping identification, load identification and modal superposition method. The error between calculated value and actual value of structure acceleration response is minimized by optimizing structural loss factors with Genetic Algorithm (GA).Then, the accurate modal damping ratio and exciting forces can be identified to optimize the parameters of acoustic prediction model. The results of comparison between prediction effect of traditional method and method presented in this paper show that, this method can improve the model prediction precision by accurately identifying the system modal damping ratio and exciting forces.