基于辅助粒子滤波的疲劳裂纹扩展预测研究

Prediction of fatigue crack propagation based on auxiliary particle filtering

针对粒子滤波算法在实现疲劳裂纹寿命预测时,存在的多样性匮乏现象而导致的精度下降问题,提出了一种基于辅助粒子滤波与结构健康监测相结合的疲劳裂纹扩展预测方法。首先将裂纹扩展的Paris规则与有限元法相结合,建立了裂纹扩展的状态方程;其次通过主动Lamb波健康监测方法,利用时间延迟损伤因子对Lamb波信号进行处理,拟合获得各裂纹长度下对应的损伤因子函数关系,建立了裂纹扩展的观测方程;最后结合状态方程和观测方程所建立的裂纹扩展状态空间模型,分别利用辅助粒子滤波与标准粒子滤波算法,实现孔边裂纹扩展的寿命预测,其预测结果的对比表明,辅助粒子滤波在复杂结构的疲劳裂纹扩展预测中,结合了最新的观测值,可有效缓解粒子多样性匮乏现象,预测结果具备更高的精度,更适用于复杂结构在线寿命预测的实现。

Aiming at the lack of diversity in the particle filtering algorithm for fatigue crack life prediction,a method for prediction of fatigue crack propagation based on auxiliary particle filtering and structural health monitoring was proposed.Firstly,Paris rule of crack propagation was combined with the finite element method to build the state equation of crack propagation.Secondly,the active Lamb wave health monitoring technique was used to monitor the process of fatigue crack propagation.The time delay damage index was used to process Lamb wave signals,and then fit the function relation between crack length and damage index,the observation equation of crack propagation was established.The state space model for crack propagation was built through combining state equations and observation equations.Finally,the life prediction of hole-edge crack propagation was realized using the auxiliary particle filtering and the standard particle filtering,respectively.The comparison of the prediction results showed that the auxiliary particle filtering in fatigue crack propagation prediction of complex structures can effectively mitigate the lack of particle diversity combining the latest observation;its prediction results are more accurate;it is more applicable for realizing complex structures’online fatigue life prediction.