摘要
提出了一种基于径向基函数神经网络的汽车结构件低载强化特性确定方法。通过足够样本的训练,构造出一个有效的神经网络模型。对于给定的强化载荷和强化次数,应用构造的神经网络模型,可以确定相应的低载强化效果。经过对某汽车前梁和传动系齿轮低载强化特性的试验验证,证实了该神经网络可以准确快捷地预测不同强化载荷和强化次数下结构的强化效果。简化了计算过程,提高了计算精度,为基于低载强化的汽车产品轻量化设计带来了方便。
A new method of determining the characteristics of low-loading strengthening for automotive structural parts is presented based on radial basis function （RBF） neural network. Through enough samples training, an effective RBF neural network can be established. With the network, the effects of low-loading strengthening can be determined under given load and times of strengthening. This is verified by the tests conducted on the front frame and transmission gears of a real vehicle. The method simplifies the calculation process, increases the calculation accuracy and facilitates the lightweight design of automotive products based on low-load strengthening.
出处
《汽车工程》
EI
CSCD
北大核心
2006年第11期993-996,共4页
Automotive Engineering
基金
国家自然科学基金项目(50375101)
上海理工大学博士启动基金资助
关键词
汽车结构件
低载强化
强化特性
神经网络
Automotive structural parts, Low-load strengthening, Strengthening characteristics, Neural network
