稳健技术的汽车底盘平顺性匹配方法研究  

Study on the matching of automobile comfort based on the robust technique

作　　者：张翠仙[1] 张国胜[2] 方宗德[3] 

机构地区：[1]西北工业大学人文与经法学院,西安710129 [2]泛亚汽车技术中心有限公司,上海201201 [3]西北工业大学机电学院,西安710072

出　　处：《现代制造工程》2010年第1期81-86,38,共7页Modern Manufacturing Engineering

摘　　要：将稳健技术应用于汽车底盘的平顺性匹配研究中,由于考虑了使用中或设计中参数变化的问题,从而使汽车在使用中能够不受噪声因素的影响而保持一定的平顺性能。利用神经网络(NN)建立汽车平顺性匹配模型,并通过该NN模型计算稳健设计中所需的导数信息,此方法不但提高了优化匹配的搜索效率,而且避免了繁琐的公式推导。建立基于1/2汽车NN模型的平顺性稳健设计方案并进行优化设计,其结果与一般设计结果进行对比,表明稳健设计结果能够在外界环境变化的情况下,保持良好的性能。The robust design is applied to the matching of automobile comfort and the better robust performances are expected. In order to improve the searching efficiency for the optimization process, uses the Neural Networks （NN） to approximate the dynamic model for the matching of the automobile comfort. The NN model is used to obtain the derivatives of the comfort performances with respect to the design and noise parameters. The contrasts between the robust design results and conventional design results for the matching of the half automobile comfort are presented and the results show the performances of the automobile designed with the robust matching method can be influenced by the change of the using conditions less and get steadier performances.

关 键 词：汽车 平顺性 匹配 稳健 神经网络 

分 类 号：U463[机械工程—车辆工程]