摘要
提出了基于有限元分析、实验设计、代理模型的血管支架轴向缩短力学性能快速分析方法,分别选取响应曲面(RSM)和径向基函数神经网络(RBFANN)作为代理模型,构建了支架轴向缩短力学性能快速分析模型,即RSM分析模型和RBFANN分析模型。研究结果表明,所建立的RSM分析模型和RBFANN分析模型均能快速地分析出支架的轴向缩短率,它们的均方根误差分别为1.0974和0.2789,相比较而言,RBFANN分析模型比RSM分析模型的分析精度要更高些。因此,本文提出的支架力学性能快速分析方法可行并且可靠,大大提高了支架力学性能分析的效率。
Based on the finite element analysis,experimental design and metamodeling techniques,this paper presents a fast analysis method of stent mechanical properties. Response surface methodology( RSM) and radial basis function artificial neural network( RBFANN) was selected respectively to construct an analysis model of stent foreshortening. The results showed that the RSM and RBFANN model established herein were able to quickly analysis the stent axial shortening rate and their root mean square error were 1. 0974 and 0. 2789,respectively. Compared to RSM model,RBFANN model had higher analysis accuracy. Therefore,the fast analysis method of stent mechanical properties presented in this paper was feasible and reliable,greatly improving the efficiency of analysis of stent mechanical properties.
出处
《功能材料》
EI
CAS
CSCD
北大核心
2015年第18期18097-18099,18104,共4页
Journal of Functional Materials
基金
国家自然科学基金资助项目(51305171)
江苏省自然科学基金资助项目(BK20130525)
江苏省高校自然科学研究资助项目(13KJB460006)
关键词
血管支架
轴向缩短
力学性能
快速分析模型
vascular stent
foreshortening
mechanical properties
rapid analysis model