A Fault Prediction Method for CNC Machine Tools Based on SE-ResNet-Transformer

Author:

Wu Zhidong12,He Liansheng1,Wang Wei1,Ju Yongzhi1,Guo Qiang1

Affiliation:

1. School of Mechanical and Electrical Engineering, Qiqihar University, Qiqihar 161006, China

2. The Engineering Technology Research Center for Precision Manufacturing Equipment and Industrial Perception of Heilongjiang Province, Qiqihar 161006, China

Abstract

Aiming at the problem that predicted data do not reflect the operating status of computer numerical control (CNC) machine tools, this article proposes a new combined model based on SE-ResNet and Transformer for CNC machine tool failure prediction. Firstly, the Transformer model is utilised to build a non-linear temporal feature mapping using the attention mechanism in multidimensional data. Secondly, the predicted data are transformed into 2D features by the SE-ResNet model, which is adept at processing 2D data, and the spatial feature relationships between predicted data are captured, thus enhancing the state recognition capability. Through experiments, data involving the CNC machine tools in different states are collected to build a dataset, and the method is validated. The SE-ResNet-Transformer model can accurately predict the state of CNC machine tools with a recognition rate of 98.56%. Results prove the effectiveness of the proposed method in CNC machine tool failure prediction. The SE-ResNet-Transformer model is a promising approach for CNC machine tool failure prediction. The method shows great potential in improving the accuracy and efficiency of CNC machine tool failure prediction. Feasible methods are provided for precise control of the state of CNC machine tools.

Funder

The Basic Research Fund for State-owned Universities in Heilongjiang Province

The Collaborative Education Project for Industry–University Cooperation Supported by the Ministry of Education

The General Research Project on Higher-Education Teaching Reform in Heilongjiang Province

The Key Projects of Qiqihar City Scientific and Technological Plan

The Educational Science Research Project of Qiqihar University

Publisher

MDPI AG

Reference33 articles.

1. Analysis and Prospect of Deep Learning Technology and Its Fault Diagnosis Application;Zhang;J. Xi’an Jiaotong Univ.,2020

2. Chen, H., Zhong, K., Ran, G., and Cheng, C. (2022). Deep Learning-Based Machinery Fault Diagnostics. Machines, 10.

3. Bidirectional Current WP and CBAR Neural Network Model-Based Bearing Fault Diagnosis;Zhu;IEEE Access,2023

4. Fault diagnosis of rollingbearingbased on continuous hidden Markov model;Hao;Chin. J. Constr. Mach.,2019

5. Ahmed, H.O.A., and Nandi, A.K. (2023). Convolutional-Transformer Model with Long-Range Temporal Dependencies for Bearing Fault Diagnosis Using Vibration Signals. Machines, 11.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3