《机电工程》杂志社科技期刊核心期刊论文投稿机电一体化机械、仪表电工自动化、计算机浙江大学主办

Introduction

International Standard Serial Number:

ISSN 1001-4551

Sponsor:

Zhejiang University;

Zhejiang Machinery and Electrical Group

Edited by:

Editorial of Journal of Mechanical & Electrical Engineering

Chief Editor:

ZHAO Qun

Vice Chief Editor:

TANG ren-zhong,

LUO Xiang-yang

Tel:

86-571-87041360,87239525

Fax:

86-571-87239571

Add:

No.9 Gaoguannong,Daxue Road,Hangzhou,China

P.C:

310009

E-mail:

meem_contribute@163.com

Rotating machinery fault diagnosis method based on depth Q learning strategy

Published:2021-12-22 author:XIN Kuo, WANG Jian-guo, ZHANG Wen-xing. Browse: 1448 Check PDF documents

Rotating machinery fault diagnosis method based

on depth Q learning strategy

XIN Kuo1,2, WANG Jian-guo1,2, ZHANG Wen-xing1,2

(1. School of Mechanical Engineering, Inner Mongolia University of Science and Technology, Baotou 014010, China;

2.Inner Mongolia Kay Laboratory of Intelligent Diagnosis and Control of Mechatronic Systems, Baotou 014010, China)

Abstract: The traditional rotating fault diagnosis methods require manual extract fault features and were greatly affected by environmental noise, aiming at the problem,a fault diagnosis model based on deep Q-network reinforcement learning was proposed. First of all, a one-dimensional fault signal was used as the input of the model, and the fault type of each fault was used as an optional action of the current input. Then, the statistical-based random zeroing method was used to improve the noise immunity of the model. The fault characteristics of each fault were effectively extracted through the deep learning network, the Q value of the current state action pair was fitted, and the deep Q learning model was used to complete each fault type identification. At last, the simulation experiment was carried out through the fault simulation test bench and the bearing fault data of Western Reserve University. It was compared with traditional machine learning methods and one-dimensional convolutional neural network models to prove the excellent performance of this method under noisy environment. The results show that when the signal-to-noise ratio was -4dB, the recognition accuracy of the fault diagnosis model can reach 78%. And This method can be used for fault diagnosis of rotating machinery accurately and stably.

Key words: rotating machinery; fault diagnosis;reinforcement learning; deep learning; noisy environment; convolution neural network (CNN)

XIN Kuo, WANG Jian-guo, ZHANG Wen-xing. Rotating machinery fault diagnosis method based on depth Q learning strategy［J］.Journal of Mechanical & Electrical Engineering, 2021,38(10):1261-1268.

Advanced Search

Online Office System

Chinese Core Periodicals
Chinese Sci-tech Core Periodicals
SA, INSPEC Indexed
CSA: T Indexed
UPD:Indexed