作　　者： ; ; ; ;

机构地区： 吉林大学机械科学与工程学院

出　　处： 《哈尔滨理工大学学报》 2013年第3期95-101,共7页

摘　　要： 针对计算机辅助某齿轮箱厂检测线上工人对齿轮箱故障状态类型的诊断,以齿轮箱中齿轮、轴、轴承等主要零件为研究对象,对采集到的齿轮箱振动信号进行细致分析,在敏感特征参数提取上采用聚类分析的方法,有效地消除了冗余参数的干扰,为下一步支持向量机(SVM)模式识提高了参数训练的精确度,建立与检测线工人所定义故障类型相一致的故障状态模式识别模型,从而进一步提高了计算机辅助工人对齿轮箱智能故障诊断的准确性和诊断结论一致性. To make fault diagnosis of the status types on the gearbox in a gear box factory line, In this paper, we take gears, shafts, bearings and other major parts as the research objects, and analyse the vibration signals collected. Then we propose cluster analysis method to extract sensitive characteristic parameters. This method can effectively eliminate the interference of the redundant parameters, and improves the accuracy of the parameter training of SVM pattern recognition process, and set up fault pattern recognition model concerned with the defination of workers＇ , so the accuracy of gearbox intelligent fault diagnosis and the consistency of the diagnosis can be improved.

关 键 词： 敏感特征参数 规则提取 模式识别

领　　域： [经济管理]