基于多模态融合的轴承剩余寿命预测技术  

Predicting the Remaining Life of Bearings Based on Multi-modal Fusion

作　　者：杨佳葳 龚涛 郭丽 李红春 彭红 YANG Jiawei;GONG Tao;GUO Li;LI Hongchun;PENG Hong

机构地区：[1]湖南宇晶机器股份有限公司,湖南益阳413001

出　　处：《今日自动化》2025年第2期149-150,170,共3页Automation Today

摘　　要：文章针对轴承剩余寿命预测的重要性及现有方法的局限性,探究了基于多模态融合的预测方法,阐述了多模态数据来源及特点,分析了不同的多模态融合策略,并结合多种预测模型进行了试验验证。试验结果表明,多模态融合方法能够提高轴承剩余寿命预测的准确性和可靠性,具有较高的可行性。The article focuses on the importance of predicting the remaining life of bearings and the limitations of existing methods.It explores prediction methods based on multimodal fusion,explains the sources and characteristics of multimodal data,analyzes different multimodal fusion strategies,and conducts experimental verification by combining multiple prediction models.The experimental results show that the multimodal fusion method can improve the accuracy and reliability of predicting the remaining life of bearings,and has high feasibility.

关 键 词：轴承剩余寿命预测 多模态融合 深度学习 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]