基于相似性的装备剩余寿命预测研究

Research on Similarity-Based Remaining Useful Life Prediction of Equipment

为了更有效估计装备的剩余寿命,对基于相似性的寿命预测方法进行研究。首先介绍基于相似性的寿命预测方法的基本概念和流程,然后研究基于高斯核回归的退化轨迹提取方法,在传统欧氏距离的基础上,考虑时间范围的影响,改进相似度计算方式,最后用高斯核密度估计得到剩余寿命的区间估计值.一个数值仿真试验表明,基于相似性的方法能够利用失效历史样本对装备的剩余寿命进行有效预测.

In order to estimate the remaining useful life （RUL）of equipment more effectively,an RUL prediction method based on similarity is put forward.In this paper,the basic concept and procedure are introduced first,and then the degrading traj ectories are extracted using the Gaussian kernel regression.The traditional similarity function based on Euclidean distance is improved by considering time frame.Finally,Gaussian kernel density estimation is applied to estimate the probability of RUL.A data experiment shows that the method based on similarity can use the failure life history of samples for effective prediction.