基于神经网络的汽车磁流变减振器非参数化建模

Non-parameter modeling of automotive magnetorheological fluid shock absorber based on neural network

结合汽车用减振器的工作特点，按照国产某微型轿车后悬架的技术要求，设计了基于混合工作模式的单出杆单筒磁流变减振器，并进行了实物样品研制。根据流体力学理论，建立混合工作模式下磁流变减振器计算模型，并对磁流变减振器的阻尼力、动态响应时间及其影响因素进行了理论分析。对设计的磁流变减振器进行台架动态特性测试，试验结果表明：单出杆单筒磁流变减振器具有较大的阻尼力与和较宽的阻尼力调节范围，并且动态响应时间较短，能满足汽车半主动悬架对减振器的使用要求。

Combine the working characteristic of automotive shock absorber,design single out single list tube magnetorheological fluid shock absorber according to the technique requirement of locally made some mini sedan rear suspension,and make real sample development.Set up magnetorheological fluid shock absorber calculation model under mixed work mode,and make theoretical analysis of the damping resistance for magnetorheological fluid shock absorber,dynamic corresponding time and its effected factors.Make bench dynamic characteristic test for designed magnetorheological fluid shock absorber.The experiment result shows： single out single list tube magnetorheological fluid shock absorber has bigger damping resistance and wide damping resistance adjustment range,and shorter dynamic corresponding time,which can meet the using requirement of automotive Semi-active for shock absorber.