南京工业大学学报（自然科学版）2024，Vol.46Issue(4)：436-444,9.DOI:10.3969/j.issn.1671-7627.2024.04.008
基于BP神经网络的木芯复合材料结构疲劳寿命预测
Fatigue life prediction of wood core composite sandwich structure based on the bp neural network
摘要
Abstract
Taking the fatigue test data from three-point and four-point bending fatigue tests of wood core composite sandwich structures under various experimental conditions as samples,constructed the load level-fatigue life(S-N)curve model and the BP neural network model to predict the fatigue life of the specimens.Additionally,the BP neural network was utilized to perform a weight analysis on the factors influencing the fatigue life of the wood core composite sandwich structures.The results indicate that the exponential function model,power function model,and BP neural network model can all effectively predict the fatigue life of the wood core composite sandwich structures,with the BP neural network model demonstrating the highest predictive accuracy.The influence of the load level,homogenized grid web area ratio,and span-to-height ratio on fatigue life decreases sequentially.Configuring the lattice webs is one of the effective methods to enhance the fatigue life of wood core composite sandwich structures.关键词
木芯复合材料结构/疲劳寿命预测/S-N曲线/BP神经网络Key words
wood core composite structure/fatigue life prediction/S-N curve/BPNN分类
建筑与水利引用本文复制引用
徐子恒,史慧媛,方海,祝典,夏志远,唐柏鉴..基于BP神经网络的木芯复合材料结构疲劳寿命预测[J].南京工业大学学报（自然科学版）,2024,46(4):436-444,9.基金项目
国家自然科学基金(52108236,52208188) （52108236,52208188）
江苏省自然科学基金(BK20200986) （BK20200986）
