Abstract:Structural Health Monitoring, Ahead of Print. In predictive maintenance for rotating machinery components, constructing health indicators is crucial for improving operational efficiency and extending the equipment lifespan. However, traditional processes of building health indicators face two significant challenges: the high cost of collecting full lifecycle data and the requirement for the operating conditions of the equipment to remain consistent with those under which the indicators were developed. These limitations make health indicators expensive and time-consuming to construct and use. This article proposed a health indicator developed based on the Joint Principal Component Cumulative Empirical Distribution Model (JPCCED-HI) to overcome these challenges. This innovative method utilizes only healthy state data, combining principal component analysis with the empirical cumulative distribution function to construct a health indicator that ranges from 0 to 1, thereby eliminating the reliance on complete lifecycle data and relaxing the requirements for specific operating conditions. The effectiveness and practical benefits of JPCCED-HI are illustrated through two detailed case studies. In the first case study, we explore the impact of different parameter settings on model performance and evaluate its anti-interference capabilities in high-noise environments based on a gearbox dataset. This study demonstrates the model's robustness in maintaining accurate health assessments despite external noise. The second case study employs a publicly available bearing dataset to compare the JPCCED-HI method against other models trained on health-stage data. This analysis reveals the superior performance of JPCCED-HI in trendability and scale similarity, affirming its effectiveness in various operational conditions. These case studies not only prove the adaptability and robustness of JPCCED-HI but also highlight its potential as a scalable solution for real-time health monitoring and predictive maintenance in industrial applications.

Constructing a Reliable Health Indicator for Bearings Using Convolutional Autoencoder and Continuous Wavelet Transform

A Reliable Health Indicator for Fault Prognosis of Bearings

New Health Indicator Construction and Fault Detection Network for Rolling Bearings via Convolutional Auto-Encoder and Contrast Learning

Health indicator construction by quadratic function-based deep convolutional auto-encoder and its application into bearing RUL prediction

A novel health indicator for intelligent prediction of rolling bearing remaining useful life based on unsupervised learning model

Asymmetric-Based Residual Shrinkage Encoder Bearing Health Index Construction and Remaining Life Prediction

Unsupervised health Indicator construction by a novel degradation-trend-constrained variational autoencoder and its applications

Health indicator construction method of bearings based on Wasserstein dual-domain adversarial networks under normal data only

A new supervised multi-head self-attention autoencoder for health indicator construction and similarity-based machinery RUL prediction

Health indicator construction for degradation assessment by embedded LSTM–CNN autoencoder and growing self-organized map

A hybrid reliability assessment method based on health index construction and reliability modeling for rolling bearing

Online anomaly detection and remaining useful life prediction of rotating machinery based on cumulative summation features

Joint optimization of degradation assessment and remaining useful life prediction for bearings with temporal convolutional auto-encoder

A Deep Learning-Based Remaining Useful Life Prediction Approach for Bearings

Deep Spatiotemporal Convolutional-Neural-Network-Based Remaining Useful Life Estimation of Bearings

A novel HI construction method based on healthy-state data training for rotating machinery components

Utilizing VQ-VAE for End-to-End Health Indicator Generation in Predicting Rolling Bearing RUL

An Automated Auto-encoder Correlation-based Health-Monitoring and Prognostic Method for Machine Bearings

Prediction of Remaining Service Life of Rolling Bearings Based on Convolutional and Bidirectional Long- and Short-Term Memory Neural Networks

A convolutional neural network based degradation indicator construction and health prognosis using bidirectional long short-term memory network for rolling bearings

Vibration-Based Wear Condition Estimation of Journal Bearings Using Convolutional Autoencoders