A method and system for predicting the residual life of complex equipment based on double-depth residual LSTM

By using a method based on dual-depth residual LSTM and processing multi-dimensional time-series monitoring data with specific functions, we have achieved rapid detection of the starting point of performance degradation of complex equipment and accurate prediction of its remaining life. This solves the safety hazards of complex equipment during the degradation period and improves prediction efficiency and accuracy.

CN115618613BActive Publication Date: 2026-07-03HARBIN INST OF TECH
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HARBIN INST OF TECH
Filing Date
2022-10-21
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

In existing technologies, the performance of complex equipment deteriorates drastically after entering the decay period, making it impossible to accurately predict the remaining service life, resulting in safety hazards and low prediction efficiency.

Method used

A dual-depth residual LSTM-based approach is adopted. By acquiring multi-dimensional time-series monitoring data of historical equipment, a first-depth residual LSTM model is trained to detect the starting point of performance degradation, and a second-depth residual LSTM model is used after training to predict the remaining lifetime. The data is processed and labeled by combining the Weibull failure rate function and the piecewise function.

Benefits of technology

It enables rapid and accurate detection of the starting point of performance degradation in complex equipment and accurate prediction of remaining life, solving the problems of large prediction errors and low efficiency in existing technologies, and improving the safety and reliability of equipment operation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115618613B_ABST
    Figure CN115618613B_ABST
Patent Text Reader

Abstract

This invention proposes a method and system for predicting the remaining lifetime of complex equipment based on dual-depth residual LSTM, comprising: acquiring multi-dimensional time-series monitoring data and performance degradation (HI) curves of historical complex equipment; using a trained first-depth residual LSTM model to detect the performance degradation initiation point of the multi-dimensional time-series monitoring data of the complex equipment under test, to determine whether the complex equipment under test has entered the degradation period; if not, the detection ends; if so, the performance degradation initiation point of the complex equipment under test is obtained, and a trained second-depth residual LSTM model is used to predict the multi-dimensional time-series monitoring data of the complex equipment under test to obtain the predicted remaining lifetime of the complex equipment under test. This application effectively solves the problems of training difficulties and performance degradation of deep LSTM while retaining the time-series data processing capabilities of traditional LSTM networks, and can quickly and accurately predict the performance degradation initiation point of complex equipment and the remaining lifetime after entering the degradation period.
Need to check novelty before this filing date? Find Prior Art