基于象限面积向量的轴心轨迹特征参数定量提取  

Feature Quantitative Extraction of Axle Centre Trail Character Parameters Based on Area Vector in Quadrant

作　　者：万波[1] 付松广[2] 梁伟[1] 李戎戈[1] 

机构地区：[1]中国石油大学(北京)机电工程学院,北京102249 [2]中国石油管道科学技术研究中心,廊坊065000

出　　处：《科学技术与工程》2007年第23期5977-5981,5986,共6页Science Technology and Engineering

摘　　要：以判断转子轴心轨迹在不同状态下,各周期的重合度和在各象限内的分布均匀性为出发点,提出了基于无量纲、象限面积向量的轴心轨迹重合度和分布均匀性的定量计算方法,并加以分析和改进。分析了在不同象限划分下,不同计算方法得到的信息熵和重合度,以及重合度标准差与状态级别之间的关系曲线。最后,结合实验,分析了上述特征提取方法的正确性,以及划分块数和重合度计算方法在状态划分中的重要作用。A quantitative compute method of axle coincidence centre trails coincidence degree and distribution uniformity degree and distribution uniformity of many circles axle centre trail in Based on this method, the relation curve was gained in different quadrant division and different compute methods, which between different statuses and entropy of area vector of axle centre trail in quadrant, coincidence degree and coincidence degree standard errors. At last, through an experiment of motor sliding bearing, the correction of feature quantitative extraction this method is examined.

关 键 词：轴心轨迹 象限面积向量 象限面积向量改进熵 重合度 重合度标准差 状态关系曲线 

分 类 号：O141.4[理学—数学]