噪声与振动控制2017，Vol.37Issue(3)：168-172,5.DOI:10.3969/j.issn.1006-1355.2017.03.033
基于奇异谱相对熵与灰色绝对关联度的监测数据特征分析
Monitoring Data Feature Analysis Based on Singular Spectrum Relative Entropy and Grey Absolute Relational Grade
摘要
Abstract
A new feature extraction method combining singular value decomposition (SVD) with relative entropy and grey absolute relational grade (GARG) algorithm is proposed. With the Wuhan-Guangzhou track as an example, the PDL GPS measurement data of wheel tread wears under 4 operating conditions, such as normal condition, slight wear condition, medium wear condition and heavy wear condition, is analyzed with this method. The variation rules of vibration signals in the 4 operation conditions are recognized. Meanwhile, the singular spectrum entropy based experiment is conducted and the results are compared with those of this method. The simulation results proved that when the wheel tread gets heavily degraded, the similarity between the normal condition signal and heavy-wear state signal gets smaller. As a result, the relative entropy value gets larger, whereas the grey relational grade value gets smaller. Therefore, the two features can effectively describe the performance degradation process of the wheel tread. What's more, the relative entropy is preferable to the singular spectrum entropy in measuring the wheel tread wearing degrees.关键词
振动与波/监测数据/车轮踏面磨损/奇异谱相对熵/灰色绝对关联度Key words
vibration and wave/monitoring data/wheel tread wearing/singular spectrum relative entropy/grey absolute relational grade分类
机械制造引用本文复制引用
于萍,金炜东,陈春利..基于奇异谱相对熵与灰色绝对关联度的监测数据特征分析[J].噪声与振动控制,2017,37(3):168-172,5.基金项目
国家自然科学基金重点资助项目(61134002) （61134002）
