文档介绍：: .
北京航空航天 TSVT 的
特性，计算了传感器之间的相关性，选取作为重构精度最佳的数据输入方向，最终使用交替乘子法实现了 LRTC-
PTNN 算法。选取 NASA 提供的 PHM2008 数据集进行实验，对数据集进行标准化并在重构后进行恢复，将多个时间
序列个数相近的发动机传感器数据构建为高维张量的形式，设置了两种传感器的数据缺失场景进行实验，结果表明，
重构后数据的 RMSE 与 MAPE 值分别介于 %～%和 %～%，LRTC-PTNN 模型优于现有的基线模
型，并且在极端情况下，该模型也有较强的鲁棒性。
关键词 航空发动机；数据缺失；张量；截断 p-shrinkage 范数；交替乘子法
中图分类号 V221+.3；TB553 文献标识码：A DOI: .1001-
Aero-engine data reconstruction based on truncated p-shrinkage norm
ZHANG Hongmei , WU Jiangnan, ZHAO Yongmei, ZENG Hang, LI Quangen
（Equipment Management and Unmanned Aerial Vehicle Engineering College, Air Force Engineering University, Xi’an 710051, China ）
*E-mail: zhm_******@
Abstract Aiming at the data loss problem of aero engine sensors, a tensor reconstruction model(A tensor completion model
based on truncated p-shrinkage norm，LRTC-PTNN) based on TSVT(Tensor singular value threshold) is proposed to
reconstruct the sensor data of aircraft engines, and LRTC-PTNN uses truncation p-shrinkage norm to replace the original
tensor trace norm as the convex envelope of tensor rank, and according to the characteristics of TSVT, the correlation betwee n
sensors is calculated, and the data input direction with the best reconstruction accuracy is selected. The LRTC-PTNN
algorithm was finally implemented using the alternating multiplier method. Using the PHM2008 dataset provided by NASA
for experiments, the dataset was standardized and restored after reconstruction, and the multiple time series similar number of
engine sensor data were constructed into the form of high-dimensional tensor,