基于多传感器精细广义复合多尺度注意熵与贝叶斯网络的齿轮箱故障诊断

打开文本图片集

DOI：10.16652/j.issn.1004-373x.2026.10.027

中图分类号：TN911.7-34；TP206.3

文献标识码：A

文章编号：1004-373X（2026）10-0184-07

Gearbox fault diagnosis based on MRGCMATE and Bayesian network

Tong Zhaojing， Fan Yongkui， Wang Pengchao

（School of Electrical Engineering and Automation， Henan University of Science and Technology， Jiaozuo ， China）

Abstract： As a commonly used transmission component in industrial equipment， it is of great significance to assess the health state of gearboxes in a timely and accurate manner. In order to improve the accuracy of gearbox health state diagnosis， a gearbox fault diagnosis method combining multi-sensor refined generalized composite multi-scale attention entropy （MRGCMATE） and Bayesian network is proposed. The complete ensemble empirical mode decomposition with adaptive noise （CEEMDAN） method is used to decompose the vibration signal into multiple IMFs， and the effective components are screened by the Pearson correlation coefficient to reconstruct the sample signals. In order to solve the problem that the traditional entropy feature extraction method is greatly affected by the parameters and the information is easy to be omitted by using single-channel information， MRGCMATE is proposed for the fault feature extraction. The obtained fault features are input into the Bayesian network improved by the Parrot optimizer for the fault recognition. The experiments were conducted on the gearbox dataset from Southeast University to verify the feasibility of the method. The experimental results show that the proposed method has a higher fault recognition accuracy compared with other fault diagnosis methods.

Keywords： gearbox; fault diagnosis; CEEMDAN; multi-sensor refined generalized composite multi-scale attention entropy; Parrot optimizer; Bayesian network

0 引言

齿轮箱作为各类机械传动系统中广泛使用的传动部件，常需要在恶劣的环境中长期运行，运行过程中易出现各种故障，进而对设备运行安全造成不良影响[1]。（剩余9215字）