辽宁石油化工大学学报

辽宁石油化工大学学报 ›› 2019, Vol. 39 ›› Issue (1): 97-100.DOI: 10.3969/j.issn.1672-6952.2019.01.018

• 信息与控制工程 • 上一篇    

基于优化BP神经网络的铝箔封口检测研究

赵士龙李维军石成江   

  1. 辽宁石油化工大学 机械工程学院,辽宁 抚顺 113001
  • 收稿日期:2018-05-04 修回日期:2018-05-24 出版日期:2019-02-25 发布日期:2019-02-28
  • 通讯作者: 李维军(1977⁃),男,博士,副教授,从事状态监测与故障诊断的研究;E⁃mail:3051379@qq.com。
  • 作者简介:赵士龙(1991?),男,硕士研究生,从事过程装备检测与控制技术方面的研究;E?mail:493290375@qq.com。

The Research of Tightness Detection Method of Aluminum Foil Seal Based on Optimized BP Neural Network Algorithm

Zhao ShilongLi WeijunShi Chengjiang   

  1. School of Mechanical Engineering, Liaoning Shihua University, Fushun Liaoning 113001,China
  • Received:2018-05-04 Revised:2018-05-24 Published:2019-02-25 Online:2019-02-28

摘要: 提出一种高效率和高精度的铝箔封口完整密封性检测方法。首先,通过红外线热像仪获取铝箔封口处的热像图,然后利用MATLAB软件批量获取热像图的信息,再对其进行图像预处理。随后用遗传算法对BP神经网络进行优化,进行图像特征提取与分类识别。实验结果证明,此密封性检测方法识别率高、训练时间短,为后续的铝箔封口生产线中的密封不良产品自动筛选和剔除提供了保证。

关键词: 铝箔密封性检测,  图像预处理,  遗传算法,  BP神经网络

Abstract: A high efficiency and high precision testing method for tightness detection method of aluminum foil seal is proposed.First of all, the thermal images of aluminum foil seal are obtained by a thermal imager.Then we use MATLAB software for getting the information of thermal image in batches,and then the image preprocessing. Then BP neural network was optimized by genetic algorithm to extract and classify image features.The experimental results show that this sealing test method has high recognition rate and short training time, which provides a guarantee for automatic screening and elimination of bad sealing products in the subsequent aluminum foil sealing production line.

Key words: Aluminum foil sealing test, ; Image preprocessingr, Genetic algorithm, BP neural network

引用本文

赵士龙, 李维军, 石成江. 基于优化BP神经网络的铝箔封口检测研究[J]. 辽宁石油化工大学学报, 2019, 39(1): 97-100.

Zhao Shilong, Li Weijun, Shi Chengjiang. The Research of Tightness Detection Method of Aluminum Foil Seal Based on Optimized BP Neural Network Algorithm[J]. Journal of Liaoning Petrochemical University, 2019, 39(1): 97-100.

使用本文

0
    /   /   推荐

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://journal.lnpu.edu.cn/CN/10.3969/j.issn.1672-6952.2019.01.018

               https://journal.lnpu.edu.cn/CN/Y2019/V39/I1/97