Deep Learning-Based Residual Useful Lifetime Prediction for Assets with Uncertain Failure Modes

2024-05-09

arXiv.org

AI Summary

To address accuracy degradation in Remaining Useful Life (RUL) prediction for industrial equipment with multiple fault modes—caused by signal overlap, unlabeled historical data, and high inter-mode fault similarity—this paper proposes a dual-model framework integrating hybrid (log-)location-scale distributions with deep neural networks. The method bypasses explicit fault identification and instead models the uncertain degradation process via unsupervised or weakly supervised degradation feature learning, yielding probabilistic RUL predictions with calibrated confidence intervals. Its key innovation lies in the first synergistic embedding of interpretable probabilistic distribution modeling within a deep learning–based RUL prediction pipeline, significantly improving both predictive accuracy and uncertainty calibration. Notably, the framework demonstrates strong robustness under fault-confusion scenarios. Numerical experiments on benchmark datasets confirm its superior overall performance compared to state-of-the-art methods.

Technology Category

Machine Learning: Calibration & Uncertainty QuantificationReasoning under Uncertainty: Relational Probabilistic ModelsIntelligent Robots: State Estimation

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Industrial prognostics focuses on utilizing degradation signals to forecast and continually update the residual useful life of complex engineering systems. However, existing prognostic models for systems with multiple failure modes face several challenges in real-world applications, including overlapping degradation signals from multiple components, the presence of unlabeled historical data, and the similarity of signals across different failure modes. To tackle these issues, this research introduces two prognostic models that integrate the mixture (log)-location-scale distribution with deep learning. This integration facilitates the modeling of overlapping degradation signals, eliminates the need for explicit failure mode identification, and utilizes deep learning to capture complex nonlinear relationships between degradation signals and residual useful lifetimes. Numerical studies validate the superior performance of these proposed models compared to existing methods.

Problem

• Multi-fault Signal Overlap
• Unlabeled Historical Data
• Prediction Accuracy Limitation

Innovation

• Hybrid Distribution
• Deep Learning
• Prognostics