华侨大学学报（自然科学版）2026，Vol.47Issue(4)：432-438,7.DOI:10.11830/ISSN.1000-5013.202510017
采用小波阈值优化与改进BP神经网络的齿轮箱故障诊断方法
Gearbox Fault Diagnosis Method Using Wavelet Threshold Optimization and Improved BP Neural Network
摘要
Abstract
To address the issue of low fault identification accuracy and poor real-time performance of marine gearbox in high-noise operating environment,a fault diagnosis method based on wavelet threshold optimization denoising and an improved back-propagation(BP)neural network is proposed.Firstly,the wavelet threshold optimization denoising technology is used to accurately extract fault signal characteristics from the gearbox.BP neural network model is subsequently constructed for feature learning and fault classification.The results dem-onstrate that the proposed method can reach a fault identification accuracy of 98.47%.Compared to other in-telligent diagnostic methods,the proposed method not only improves the fault recognition accuracy,but also demonstrates better real-time performance.关键词
船用齿轮箱/小波阈值/自适应神经网络/特征学习/故障预测Key words
marine gearbox/wavelet threshold/adaptive neural network/feature learning/fault prediction分类
信息技术与安全科学引用本文复制引用
林金亮,胡新福..采用小波阈值优化与改进BP神经网络的齿轮箱故障诊断方法[J].华侨大学学报（自然科学版）,2026,47(4):432-438,7.基金项目
国家自然科学基金资助项目(52475609) （52475609）
福建省龙岩市科技计划重点项目(2022LYF9007) （2022LYF9007）
闽西职业技术学院加强校村合作、助力乡村振兴研究专项成果(MXZY25XC11) （MXZY25XC11）
