China Mechanical Engineering ›› 2021, Vol. 32 ›› Issue (07): 778-785，792.

### Rolling Bearing Fault Diagnosis Method Based on Adaptive Autogram

ZHENG Jinde1,2;WANG Xinglong1;PAN Haiyang1;TONG Jinyu1;LIU Qingyun1

1. 1.School of Mechanical Engineering,Anhui University of Technology,Maanshan,Anhui,243032
2. Anhui Key Laboratory of Mine Intelligent Equipment and Technology,Anhui University of Science & Technology,Huainan,Anhui,232001
• Online:2021-04-10 Published:2021-04-16

### 基于自适应自相关谱峭度图的滚动轴承故障诊断方法

1. 1.安徽工业大学机械工程学院，马鞍山，243032
2.安徽理工大学矿山智能装备与技术安徽省重点实验室，淮南，232001
• 作者简介:郑近德，男，1986年生，副教授。研究方向为设备状态监测与故障诊断、非平稳信号处理、模式识别、非线性动力学。E-mail:lqdlzheng@126.com。
• 基金资助:
国家重点研发计划（2017YFC0805100）；
国家自然科学基金（51975004）；
安徽省高校自然科学研究重点项目（KJ2019A0053,KJ2019A092）；
安徽理工大学矿山智能装备与技术安徽省重点实验室开放基金(201902005)

Abstract: In Autogram method, the signal spectrum was divided by the maximum overlap discrete wavelet packet transform, and the signals in the frequency band corresponding to the maximum kurtosis value were selected for diagnostic analysis. However, the method followed the binary tree structure when the frequency band was divided, and the division area of this structure was fixed. A fault diagnosis method of rolling bearings was proposed based on adaptive Autogram to solve this problem. The improved empirical wavelet transform was used as the basis of adaptive Autogram. In this process, the original signal Fourier spectrum was enveloped and smoothed and then segmented, thus achieving the purpose of frequency band was adaptively divided by Autogram. The simulation signals and experimental data were analyzed through the proposed method, and the analysis results were compared with the existing fast kurtogram and Autogram. The results show that the optimal demodulation frequency band may be accurately detected by the proposed method, and the fault characteristics are more obvious.

CLC Number: