基于Bayes定理和μGA的结构损伤识别方法研究

Study on Structural Damage Identification Using Bayes' Theorem and μGA

提出了一种基于Bayes定理和μGA(微种群遗传算法)的方法,用于识别结构中损伤的位置和程度等参数并分析其不确定性。该方法在获得测量信息后,结合损伤模型,根据Bayes定理更新获得损伤模型参数的后验概率密度函数,以实现在不确定性情况下对损伤的识别。为最大化损伤模型参数的后验概率密度函数,采用μGA搜索获得描述损伤的全局最优参数。将该方法应用于板结构损伤识别,并进行了数值仿真研究以验证所提方法的有效性。

This paper proposed an approach for estimating the parameters of the location and size of damage in structure along with the associated uncertainties using Bayes＇ Theorem and micro genetic algorithm（μGA）.After obtaining the measured information,combined with the damage model,the posterior probability density function（PDF） of the para-meters of the damage was updated according to the Bayes＇ Theorem,thus the damage was identified with uncertainties.To maximize the updated PDF of damage parameters,μGA was employed to search out the global optimum values of damage parameters.The proposed approach was applied to damage identification for plate structure,and numerical study was performed to demonstrate its effectiveness.