Convolutional Neural Networks—Long Short-Term Memory—Attention: A Novel Model for Wear State Prediction Based on Oil Monitoring Data

Wear state prediction based on oil monitoring technology enables the early identification of potential wear and failure risks of friction pairs, facilitating optimized equipment maintenance and extended service life. However, the complexity of lubricating oil monitoring data often poses challenges i...

Full description

Saved in:
Bibliographic Details
Main Authors: Ying Du, Hui Wei, Tao Shao, Shishuai Chen, Jianlei Wang, Chunguo Zhou, Yanchao Zhang
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Lubricants
Subjects:
Online Access:https://www.mdpi.com/2075-4442/13/7/306
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Wear state prediction based on oil monitoring technology enables the early identification of potential wear and failure risks of friction pairs, facilitating optimized equipment maintenance and extended service life. However, the complexity of lubricating oil monitoring data often poses challenges in extracting discriminative features, limiting the accuracy of wear state prediction. To address this, a CNN–LSTM–Attention network is specially constructed for predicting wear state, which hierarchically integrates convolutional neural networks (CNNs) for spatial feature extraction, long short-term memory (LSTM) networks for temporal dynamics modeling, and self-attention mechanisms for adaptive feature refinement. The proposed architecture implements a three-stage computational pipeline. Initially, the CNN performs hierarchical extraction of localized patterns from multi-sensor tribological signals. Subsequently, the self-attention mechanism conducts adaptive recalibration of feature saliency, prioritizing diagnostically critical feature channels. Ultimately, bidirectional LSTM establishes cross-cyclic temporal dependencies, enabling cascaded fully connected layers with Gaussian activation to generate probabilistic wear state estimations. Experimental results demonstrate that the proposed model not only achieves superior predictive accuracy but also exhibits robust stability, offering a reliable solution for condition monitoring and predictive maintenance in industrial applications.
ISSN:2075-4442