衬套模型对商用车驾驶室载荷分解精度的影响研究  

Study on Influence of Bushing Model on Load Decomposition Accuracy of Commercial Vehicle Cab

作　　者：芦伟[1] 李伟[1] 张艳玲[1] Lu Wei;Li Wei;Zhang Yanling(Anhui Jianghuai Automobile Co.,Ltd.,Hefei 230601)

机构地区：[1]安徽江淮汽车股份有限公司,合肥230601

出　　处：《汽车技术》2020年第7期35-39,共5页Automobile Technology

摘　　要：针对商用车驾驶室疲劳载荷分解过程中衬套建模精度不足的问题,研究了基于神经网络和样条插值的衬套建模方法,并与衬套刚度试验结果进行对比,结果表明,基于神经网络的衬套模型在随机波形试验数据测试集上的精度提升较样条插值模型更明显。基于以上两种衬套模型分别建立驾驶室多体动力学模型,采用虚拟迭代法提取驾驶室疲劳载荷,在短波路工况上进行验证,发现基于神经网络的衬套建模方法的载荷分解精度较基于样条插值的建模方法提高了8.41%,且两种衬套建模方法都满足工程需要。To solve the problem of insufficient modeling accuracy of the bushing during fatigue load decomposition of the commercial vehicle cab,two kinds of bushing modeling methods are studied based on neural network and spline interpolation,and compared with the bushing stiffness test results.The results show that the bushing model based on neural network is more accurate than the spline interpolation model in the experimental data test set of random wave forms.Based on the above two bushing models,the multi-body dynamics model of the cab is established respectively,and the fatigue load of the cab is extracted by using virtual iteration.With the short-wave path as the verification condition,the loaddecomposition accuracy of the bushing modeling method based on neural network is 8.41%higher than that based on spline interpolation,and both bushing modeling methods meet engineering requirements.

关 键 词：神经网络 样条插值 衬套建模 载荷分解 

分 类 号：U461.71[机械工程—车辆工程]