Abstract:Catering to the rapidly growing of smart manufacturing, recent years have witnessed that a great number of degradation trend prediction techniques were developed with the desire to guarantee the reliability and safety of machinery. Most existing studies assume that the future degradation trends follow similar data distributions to the degradation features fed for modeling, which may be challenging to satisfy in realistic scenarios. To address this issue, distribution discrepancy aligning-oriented domain adaptation was proposed. However, the stage-specific properties of the degradation process are often ignored, which may not be conducive. to maintain consistency in the distribution within the degradation stage and to evaluate the temporal distribution discrepancy in the degradation process. Therefore, this study presents a performance degradation trend prediction methodology that incorporates considerations of the temporal distribution discrepancy between degradation stages to overcome these challenges. Specifically, an adaptive degradation stage characterization (AdaDSC) strategy is designed to partition the degradation feature sequence into multiple degradation stages, ensuring that their distributions are as diverse as possible. Building upon this, a distribution discrepancy evaluation gated recurrent unit (DDE-GRU) module is introduced. It aims to mitigate the effects of distribution gaps while capturing the temporal dependencies within the data. The feasibility and effectiveness of the proposed methodology are verified through experiments conducted on three publicly available run-to-failure datasets, as well as an engineering dataset collected from the main bearing of a wind turbine. These experiments demonstrate the potential and practicality of the proposed methodology in accurately predicting performance degradation trends.

Self-adaptive Health Condition Prediction Considering Dynamic Transfer of Degradation Mode

An adaptive method for health trend prediction of rotating bearings.

Multiple Degradation Mode Analysis Via Gated Recurrent Unit Mode Recognizer and Life Predictors for Complex Equipment

Health Trend Prediction of Rolling Bearings using Transferable Autoregressive Recurrent Adaptation Approach

Bearing Performance Degradation Assessment and Remaining Useful Life Prediction Based on Data-Driven and Physical Model

Deep transfer learning-based hierarchical adaptive remaining useful life prediction of bearings considering the correlation of multistage degradation

An Efficient Sequential Method for Predictive Maintenance of Equipment Health

An Autoregressive Model-Based Degradation Trend Prognosis Considering Health Indicators with Multiscale Attention Information

An Integrated Method for Bearing State Change Identification and Prognostics Based on Improved Relevance Vector Machine and Degradation Model

Machinery Degradation Trend Prediction Considering Temporal Distribution Discrepancy Between Degradation Stages

Data-Driven Performance Assessment And Prediction Approach For Machinery Prognostics

An Adaptive Method for Predicting Bearing Remaining Useful Life Across Various Degradation Stages

Two-stage prediction technique for rolling bearings based on adaptive prediction model

An Adaptation-Aware Interactive Learning Approach for Multiple Operational Condition-Based Degradation Modeling

Bearing Degradation Prediction Based on Deep Latent Variable State Space Model with Differential Transformation

A Two-Stage Data-Driven-Based Prognostic Approach for Bearing Degradation Problem.

A Hybrid Degradation Modeling and Prognostic Method for the Multi-Modal System

Exponential Smoothing Model for Condition Monitoring: A Case Study

Data-driven Dynamic Predictive Maintenance for a Manufacturing System with Quality Deterioration and Online Sensors.

Health Status-Based Predictive Maintenance Decision-Making via LSTM and Markov Decision Process

Deep transfer learning based on dynamic domain adaptation for remaining useful life prediction under different working conditions