基于多尺度模糊熵的齿轮故障诊断方法  

Gear Fault Diagnosis Method Based on Multiscale Fuzzy Entropy

作　　者：吴英建 王景霖 刘贞报[2] WU Ying-jian;WANG Jing-lin;LIU Zhen-bao(Aviation Key Laboratory of Science and Technology on Fault Diagnosis and Health Management,AVIC Shanghai Aero Measurement-Controlling Research Institute,Shanghai 201601,China;Northwestern Polytechnical University,Xi'an 710072,China)

机构地区：[1]上海航空测控技术研究所故障诊断与健康管理技术航空科技重点实验室,上海201601 [2]西北工业大学,陕西西安710072

出　　处：《测控技术》2021年第3期19-23,共5页Measurement & Control Technology

基　　金：航空科学基金重点项目(20173333002)。

摘　　要：齿轮作为旋转机械的关键零部件之一,其健康状态会影响机械设备的正常运行,因此需要对齿轮进行故障诊断。为了克服模糊熵从单一尺度上考虑时间序列复杂度不够全面的问题,采用了多尺度模糊熵从多个尺度对信号进行处理从而提取故障特征,并借助对类域的交叉或重叠较多的待分样本集识别效果显著的K最近邻分类器对提取的多尺度模糊熵特征进行分类,确定齿轮是否发生故障和发生故障的类型。为了验证提出方法的有效性,使用齿轮故障试验台采集相关数据集对方法进行测试并与多尺度熵以及根据时间和频率特性提取的特征进行对比,提出的方法对5种不同的齿轮故障类型识别率达到了100%,明显优于两种对比特征提取方法,为齿轮故障诊断提供了新思路。Gear is one of the key parts of rotating machinery,its health status will affect the normal operation of mechanical equipment,so it is necessary to carry out fault diagnosis for gear.In order to overcome the problem that the complexity of time series from a single scale is not comprehensive enough,multi-scale fuzzy entropy is used to process the signal from multiple scales to extract fault features,and K-nearest neighbor classifier is used to classify the extracted multi-scale fuzzy entropy features to determine whether the gear fault accurs and the type of falut.In order to verify the effectiveness of the proposed method,the gear fault test-bed is used to collect relevant data sets to test the method,and compared with the multi-scale entropy and the features extracted according to the time and frequency characteristics.The recognition rate of the proposed method for five different gear fault types reaches 100%,which is significantly better than the two comparative feature extraction methods.It provides a new method for gear fault diagnosis ideas.

关 键 词：故障诊断 特征提取 多尺度模糊熵 齿轮 

分 类 号：TH165.3[机械工程—机械制造及自动化]