摘要
为了提升发动机罩盖的NVH性能,针对振型为扭转和弯曲的两种情况,研究设计参数对模态的影响,选取罩盖设计中的关键参数为变量,运用最优拉丁超立方技术进行DoE分析;基于RBF神经网络方法搭建模态和变量之间的近似模型,建立模态对变量的响应关系。结果表明,不同变量对同一振型模态的贡献量、相关性及响应关系不同,且在两种振型下的结果有明显差异,为车型开发提供了有效参考。
In order to improve the NVH performance of engine hood and study the relationships of design parameters to modal under torsion or bending vibration mode,some key design parameters are selected as variables.The DoE analysis is developed by Optimal Latin Hypercube Sampling method.The approximation model and response relationship between modal and variables are established using RBF neural network technology.The results show that the contribution,correlation and response relationship of different variables to modal under the same vibration mode are different.And the difference under the two vibration modes is obvious.The conclusion is effective to support the development of products.
作者
曹广军
潘月华
朱敏
CAO Guangjun;PAN Yuehua;ZHU Min
出处
《上海汽车》
2020年第2期25-28,38,共5页
Shanghai Auto