摘要
针对数学形态谱熵难以准确描述信号的形态复杂度以及性能退化趋势评价效果不理想的问题,提出一种基于数学形态梯度谱熵的性能退化特征提取方法。该方法利用形态梯度算子在信号处理中能有效提取故障特征信息并抑制噪声的优势,将其引入到形态谱熵的定义中,得到数学形态梯度谱熵的概念。通过对仿真信号进行分析,验证了所提出的形态梯度谱熵作为信号复杂度指标的合理性与有效性。最后,将该方法应用到滚动轴承的性能退化研究中,结果表明,形态梯度谱熵能有效反映滚动轴承的性能退化趋势。
Aiming at that morphology spectrum entropy fails to accurately describe the morphological complexity of signals and its evaluation effect of performance degradation trend is not ideal,a performance degradation feature extraction method based on mathematical morphological gradient spectrum entropy was proposed in this work.Morphological gradient algorithm which could effectively extract fault feature information and remove interference components in the signal processing was introduced into the definition of morphology spectrum entropy.The concept of mathematical morphological gradient spectrum entropy was thus obtained.The simulations verify the rationality and effectiveness of the morphological gradient spectrum entropy as a signal complexity index.Lastly the rolling bearing performance degradation study results prove that the morphological gradient spectrum entropy can describe rolling bearing＇s performance degradation trend.
出处
《振动与冲击》
EI
CSCD
北大核心
2016年第16期86-90,共5页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(51275524)
关键词
性能退化
特征提取
形态谱熵
形态梯度谱熵
滚动轴承
performance degradation
feature extraction
morphological spectrum entropy
morphological gradient spectrum entropy
rolling bearing
