改进HPO优化VMD-GRU法在刀具磨损状态识别中的应用

韩宁; 李国富; 任潞

doi:10.13433/j.cnki.1003-8728.20230357

Chinese

您当前的位置：

首页 >

文章列表页 >

改进HPO优化VMD-GRU法在刀具磨损状态识别中的应用

更新时间：2026-04-23

- 改进HPO优化VMD-GRU法在刀具磨损状态识别中的应用
- Application of Improved HPO Optimized VMD and GRU Method in Tool Wear State Recognition
- 机械科学与技术 2025年44卷第11期页码：1912-1918
- 作者机构：
  
  宁波大学机构机械工程与力学学院,浙江,宁波,315211
- 作者简介：
  
  [ "韩宁，硕士研究生，" ]
  [ "李国富，教授，硕士生导师，博士，" ]
- 基金信息：
  
  国家自然科学基金项目（22108316）
- DOI：10.13433/j.cnki.1003-8728.20230357
  中图分类号：
- 纸质出版：2025
- 稿件说明：
移动端阅览
韩宁, 李国富, 任潞. 改进HPO优化VMD-GRU法在刀具磨损状态识别中的应用[J]. 机械科学与技术, 2025,44(11):1912-1918.

韩宁, 李国富, 任潞. Application of Improved HPO Optimized VMD and GRU Method in Tool Wear State Recognition[J]. Mechanical Science and Technology for Aerospace Engineering, 2025, 44(11): 1912-1918.
韩宁, 李国富, 任潞. 改进HPO优化VMD-GRU法在刀具磨损状态识别中的应用[J]. 机械科学与技术, 2025,44(11):1912-1918. DOI： 10.13433/j.cnki.1003-8728.20230357.

韩宁, 李国富, 任潞. Application of Improved HPO Optimized VMD and GRU Method in Tool Wear State Recognition[J]. Mechanical Science and Technology for Aerospace Engineering, 2025, 44(11): 1912-1918. DOI： 10.13433/j.cnki.1003-8728.20230357.

摘要

为提高车刀在加工过程中的磨损状态识别精度，提出了一种基于改进猎人猎物算法(Hunter-prey optimizer

HPO)、变分模态分解（Variational mode decomposition

VMD）及门控循环单元(Gated Recurrent Unit

GRU)的识别模型。首先，采用Tent混沌和Levy飞行策略改进的猎人猎物算法优化VMD确定分解层数和惩罚因子的最佳组合。其次，对原始电流信号及分解后的信号进行特征提取，并通过核主成分分析法（Kernel principal component analysis

KPCA）进行特征降维。最后，将降维后的特征输入GRU神经网络模型实现刀具磨损状态的识别。实验结果表明，所提出的识别模型与其他模型相比具有更高的识别精度和效率，且具有较高的泛用性。

Abstract

In order to improve the accuracy of the tool wear state identification during machining

a recognition model based on improved Hunter-prey optimizer (HPO)

variational mode decomposition (VMD) and gated recurrent unit (GRU) neural network is proposed in this paper. Firstly

the HPO improved by Tent chaos and Levy flight strategy was used to optimize VMD to determine the best combination of the decomposition layers and the penalty factor. Then

the features of the original current signal and the decomposed signal are extracted and the kernel principal component analysis (KPCA) method is used to reduce the feature dimension. Finally

the dimensionality reduction features were input into the Gated Recurrent Unit (GRU) neural network model to realize the recognition of tool wear states. The experimental results show that the recognition model proposed in this paper has higher recognition accuracy and efficiency than other models

and has higher universality.

关键词

Keywords

references

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

变分模态分解和自适应稀疏自编码器的故障诊断模型

电流时域同步平均残差特征统计的滚珠丝杠故障定位方法研究

一种北方苍鹰参数优化的VMD-MRE轴承故障诊断方法

矿用对旋风机旋转失速信号特征识别与提取