融合应力集中因子与贝叶斯理论的结构健康评估方法

针对传统结构健康评估方法评估精度较低的现状，提出一种融合应力集中因子与贝叶斯理论的结构健康评估方法。首先，结合基于应力集中因子的序化模型策略与非齐次泊松过程，运用贝叶斯理论方法对试验信息及总体过程参数的序化关系进行有效融合，推导出适合不同初始损伤状态的过程参数先验分布。同时，基于应力集中递进因子与似然函数的概念，建立融合应力集中因子与贝叶斯定理非齐次泊松过程的参数后验分布预测方法。最后对提出的结构健康评估方法进行实例研究，并将预测结果与独立的试验结果进行比较。结果表明，提出方法的平均预测精度为91.99%，可以实现通过少量试验信息完成结构健康状态评估的目的，从而为结构健康的精确评估提供了一种思路。

Health Assessment Approach for Structure by Fusion of Stress Concentration Factor and Bayesian Theory

Aiming at the low accuracy of traditional structural health assessment approach, the health assessment approach for structure by fusion of stress concentration factor and Bayesian theory is proposed. The prior parameters distribution for different damage situation based on non-homogeneous Poisson process is obtained by combine the ordering model of stress concentration factor with the non-homogeneous Poisson model, and having an effective integration on the test information with the ordering relation, by using Bayesian method. The posterior distribution calculation approach based on Bayesian is achieved by using the prior information, based on the progressive factor and the likelihood function. Further, a case study of the proposed structural health assessment approach is carried out, and the predicted results are compared with the independent test results. It can be clearly found from the results that the average accuracy is 91.99%. The proposed approach can be used to completes the purpose of health assessment by using a small amount of test information, and provides a way for the accurate assessment of structural health.