摘要
现有针对航空航天设备、电子器件等高可靠度产品的寿命预测方法主要研究如何建立寿命预测模型,进而预测产品的寿命,缺少对于模型的验证及准确度评估。然而,模型是否准确会直接影响寿命预测的精度。另外,同一组数据,采用不同的寿命预测方法,得到的预测寿命往往相差较大。如何选择合适的寿命预测模型是值得研究的问题。针对以上问题,本文提出基于交叉验证的加速寿命预测模型评估方法,为模型的准确度评估以及模型的选择提供了一种新的思路。在基于随机过程的加速寿命预测、基于退化轨迹的加速寿命预测、基于退化量分布的加速寿命预测实验中,利用本文方法对各寿命预测模型进行评估,评估结果作为模型选择的依据,验证了本文方法的有效性。
The existing lifetime prediction methods for aerospace equipment,electronic devices and other high reliability products mainly study how to establish a lifetime prediction model to predict the life of products,but lack of verification and accuracy evaluation of the model.However,the accuracy of the model will directly affect the accuracy of lifetime prediction.In addition,the predicted life of the same set of data often varies greatly with different lifetime prediction methods.How to choose an appropriate life prediction model is a problem worthy of study.In view of the above problems,this paper proposes an accelerated life prediction model evaluation method based on cross validation,which provides a new idea for the accuracy evaluation of the model and the selection of the model.In the experiments of accelerated lifetime prediction based on random process,accelerated lifetime prediction based on degradation trajectory and accelerated lifetime prediction based on degradation quantity distribution,this method is used to evaluate each lifetime prediction model.The evaluation results are used as the basis for model selection to verify the effectiveness of this method.
作者
龚丹丹
姜昊
袁颖
GONG Dandan;JIANG Hao;YUAN Ying(School of Statistics and Information,Shanghai University of International Business and Economics,Shanghai 201620,China)
出处
《智能计算机与应用》
2024年第9期10-18,共9页
Intelligent Computer and Applications
基金
上海市哲学社会科学规划一般资助项目(2019BGL003)。
关键词
航天设备
电子器件
寿命预测
模型评估
交叉验证
space equipment
electronic device
lifetime prediction
model evaluation
cross validation
