Integration of symbolic regression and domain knowledge for interpretable modeling of remaining fatigue life under multistep loading
Integration of symbolic regression and domain knowledge for interpretable modeling of remaining fatigue life under multistep loading
复制标题
DOI：
10.1016/j.ijfatigue.2022.106889
复制
发表时间：
2022-04-07
影响因子：
6
通讯作者：
Zhong, Zheng
中科院分区：
文献类型：
--
作者：
Gan, Lei;Wu, Hao;Zhong, Zheng
本研究旨在探索将数据驱动的符号回归（SR）与领域知识相结合，建立多步载荷下的剩余疲劳寿命模型。为此，对6种经典的半经验损伤模型进行了分析，提取出可靠的领域知识作为对SR公式结构的约束。同时，收集了15种材料和结构、3种载荷谱共194个实验结果作为数据支撑。成功地建立了一种新的不含拟合参数的两步加载剩余疲劳寿命估计模型，这是本研究工作的主要贡献。该模型可以在常规损伤模型的框架内进行解释，并通过对损伤指标和损伤过渡的适当定义，对多步加载具有良好的可扩展性。大量的模型评估表明，该模型在预测精度和适用范围上都优于现有的5种损伤模型，对多步载荷下的剩余寿命估计具有较强的适用性。
This research work aims to explore the integration of data-driven symbolic regression (SR) and domain knowledge to model the remaining fatigue life under multistep loading. To this end, six classical semiempirical damage models are analyzed to distill reliable domain knowledge as the restrictions on the structures of SR formulas. Meanwhile, a total of 194 experimental results involving fifteen materials and structures as well as three kinds of loading spectrums are collected for data support. As a major contribution of this research work, a novel model without including fitting parameters is successfully discovered for remaining fatigue life estimation under two-step loading. This model can be interpreted in the framework of conventional damage models, and shows good extendibility to multistep loading through proper definitions of the damage indicator and the damage transition. Extensive model evaluations demonstrate that the discovered model is better than five existing damage models in terms of predictive accuracy and application scope, showing great applicability for remaining life estimation under multistep loading.