神经网络法预测汽车轮胎的微观与宏观性能

发布时间：2018-02-11 04:28

  本文关键词： 设计参数 有限元模拟 神经网络 轮胎构型　出处：《现代橡胶技术》2016年05期 　论文类型：期刊论文

【摘要】：有限元(FE)分析已成为轮胎行业虚拟研究轮胎备受青睐的工具,因为它能模拟轮胎胎体的接合部细节。然而,在轮胎设计开发中应用有限元分析依然非常耗时,且花费不菲。在此,对应用各种人工神经网络(ANN)结构来预测轮胎性能进行了评估,以便选择最有效和最高效的结构。这样我们可在用花费高得多的全过程有限元分析进行证实预测的性能之前,以花费不多的费用进行广泛的参数研究,以便优化轮胎设计。
[Abstract]:Finite element (FEE) analysis has become a popular tool in tire virtual research because it can simulate the joint details of tire carcass. However, the application of finite element analysis in tire design and development is still time-consuming. And it's expensive. Here, we evaluate the performance of tires using a variety of artificial neural network (Ann) structures. In order to select the most effective and efficient structure, we can carry out extensive parametric studies at a low cost to optimize the tire design before verifying the predicted performance with a much more expensive finite element analysis of the whole process.
【分类号】：U463.341;TP183
，

本文编号：1502204