欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出将奇异值分解（SVD）与相对熵、灰色绝对关联度算法相结合，对武广线车轮踏面磨损4工况监测数据提取特征进行分析，旨在充分认识踏面工作状态经历正常-轻微磨损-中度磨损-严重磨损接近镟修这一过程中的信号变化规律，并开展基于奇异谱熵特征提取的对照实验。后续仿真结果表明：踏面性能退化越深，其监测信号与正常状态的相似性就越小，所得奇异谱相对熵特征值越大，灰色绝对关联度特征值就越小，即此二维两个特征是衡量车轮踏面性能退化过程的有效指标；其次，奇异谱相对熵的特征分析结果明显优于对照实验中的奇异谱熵。
英文摘要：
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 under4operating 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 the4operation 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.
同期刊论文项目
同项目期刊论文
期刊信息
位置：