基于多源信息融合的飞行器部件剩余寿命预测

针对飞行器关键部件的多源变量数据统计信息,提出基于多源信息融合的相似性剩余寿命预测方法。介绍了相似性剩余寿命预测方法的基本思想和模型;提出一种使用BP神经网络融合多变量统计数据的方法;引入余弦相似度方法,将服役部件和参考部件退化模型进行模式匹配,确定与服役部件具有相同退化模式的参考部件,进而提高基于相似性剩余寿命预测方法的预测精度。通过NASA航空发动机数据集和相同评价指标下的对比分析,验证了该方法的有效性。

Residual Life Prediction of Aircraft Components Based on Multi-source Information Fusion

For the multi-source variable data statistical information of key components of aircraft,a similarity-based remaining life prediction method based on multi-source information fusion is proposed.The basic idea and model for the remaining life prediction are introduced,a method of using BP neural network to fuse multivariate statistical data is proposed.Then,the cosine similarity method is introduced to match the degradation models of service components and reference components,thus determining the reference components with the same degradation mode as the service components,so as to improve the prediction accuracy of similarity-based remaining life prediction method.The validity of this method is verified by the NASA aeroengine data set and comparison analysis under the same evaluation index.