基于改进激活函数的一维卷积神经网络电机轴承故障诊断的研究

Research on fault diagnosis of motor bearings based on one-dimensional convolutional neural network based on improved activation function

任大卫 ¹周舒昊 ¹伦淑娴 ¹李明¹

扫码查看

作者信息

1. 渤海大学控制科学与工程学院,辽宁锦州 121013
折叠

摘要

提出了一种基于改进激活函数的一维卷积神经网络的电机轴承故障诊断的方法,该方法首先介绍了一维卷积神经网络的结构,然后详细说明了激活函数的改进点,最后通过仿真试验依次采用三种一维卷积神经网络对电机轴承故障进行分类,通过对比发现,此方法具有诊断准确率高、收敛速度快、无需人为提取故障特征等优点.

Abstract

This paper introduces a fault diagnosis method for motor bearings based on a one-dimensional convolutional neural network with an improved activation function.Firstly,the structure of one-dimensional convolutional neural network is introduced.Then the improvement of the activation function is introduced.Finally,three kinds of one-dimensional convolutional neural networks are used to classify the faults of motor bearings through simulation experiments.

关键词

激活函数/一维卷积神经网络/电机轴承故障诊断

Key words

activation function/one-dimensional convolutional neural network/motor bearing fault diag-nosis

引用本文复制引用

出版年

2024

渤海大学学报(自然科学版)

渤海大学

渤海大学学报(自然科学版)

影响因子：0.37

ISSN：1673-0569

参考文献量3

段落导航