Aero-engine residual life prediction method based on transfer learning

Abstract

An aero-engine residual life prediction method based on transfer learning comprises the following steps: performing sensitivity analysis on multi-dimensional degradation parameters of an aero-engine, screening out degradation sensitive parameters, and preprocessing the degradation sensitive parameters to improve data expression ability; carrying out mobility measurement on a to-be-tested sample based on a dynamic time warping algorithm, and finishing time domain sensitive data matching by utilizing a sliding time window; on the basis of a mobility measurement result, screening a migratable sample and calculating a migration weight; constructing a training data set by cutting migratable sample data; training a plurality of residual life prediction models by using the training data of each transferable sample; and realizing prediction result fusion of the life prediction models based on the transferable weight to obtain a final prediction result.

Description

technical field [0001] The invention relates to the technical field of prediction of the remaining life of an aero-engine, in particular to a method for predicting the remaining life of an aero-engine based on transfer learning. Background technique [0002] The aero-engine is the most important part of the aircraft. Due to its complex structure and harsh working environment, the failure modes of the aero-engine are more diverse and more prone to failure than other parts of the aircraft. Accurately predicting the remaining service life of aero-engines is of great significance for improving the safety and reliability of aero-engines and avoiding the waste of maintenance resources caused by excessive maintenance. [0003] Existing aero-engine remaining service life prediction methods can be divided into model-based methods, statistical analysis-based methods, and data-driven methods. Although the model-based prediction method can more accurately reflect the degradation mechan...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to disclose a method for predicting the remaining life of an aero-engine based on transfer learning. The technical problem solved by this method is that the current data-driven prediction method is difficult to obtain degradation data, poor engineering usability, and low prediction accuracy.

Images