机电工程技术2025，Vol.54Issue(14)：58-62,103,6.DOI:10.3969/j.issn.1009-9492.2025.14.010
基于信号相似度动态评估的滚动轴承运行状态识别方法研究
Research on Rolling Bearing Running State Recognition Based on Dynamic Evaluation of Signal Similarity
摘要
Abstract
To enhance the accuracy of monitoring the operating condition of rolling bearings,a recognition method based on signal similarity dynamic evaluation is proposed.The method addresses issues in traditional approaches,such as the impact of phase fluctuations,insufficient spectral features,and low fault warning accuracy.By reconstructing the frequency domain features of vibration signals using Gram angular field,more representative non-redundant spectra is generated,thus improving the expressiveness of signal features.On this basis,the Frobenius norm between the test signal and the reference signal is calculated to quantify the signal similarity differences in the Gram angular field matrix,with a threshold set to monitor changes in the rolling bearing's condition in real-time.Experimental results show that the method achieves high accuracy in state recognition while reducing the average warning time by 20%,demonstrating strong practicality.Compared to existing methods,the proposed approach offers significant advantages in improving diagnostic accuracy and real-time performance,particularly excelling in early fault detection and preventive maintenance,providing a more practical and innovative solution for identifying the operating condition of rolling bearings.关键词
频域相似度/格拉姆角场/滚动轴承/状态监测Key words
frequency domain similarity/Gram angle field/rolling bearing/condition monitoring分类
机械制造引用本文复制引用
陈振洋,戚雪峰,王淑华..基于信号相似度动态评估的滚动轴承运行状态识别方法研究[J].机电工程技术,2025,54(14):58-62,103,6.基金项目
浙江省教育厅科研项目(Y202249547) （Y202249547）
