吉林大学学报(信息科学版)

• 论文 • 上一篇    下一篇

基于改进的卷积神经网络的钢号识别

任伟建1a,1b ,宋 月1a,1b ,陈建玲2 ,任 璐3 ,孙勤江4   

  1.  1. 东北石油大学 a. 电气信息工程学院; b. 黑龙江省网络化与智能控制重点实验室,黑龙江 大庆 163318;2. 中海石油( 中国) 有限公司天津分公司 渤海研究院,天津 300452;3. 海洋石油工程股份有限公司 设计公司,天津 300450; 4. 中海石油( 中国) 有限公司 天津分公司,天津 300452
  • 收稿日期:2017-05-06 出版日期:2018-05-24 发布日期:2018-07-25
  • 作者简介:任伟建( 1963— ) ,女,黑龙江泰来人,东北石油大学教授,博士生导师,主要从事复杂系统的建模与控制研究,( Tel)86-13845901386( E-mail) renwj@ 126. com。
  • 基金资助:
    国家自然科学基金资助项目( 61374127) ; 黑龙江省博士后科研启动资金资助项目( LBH-Q12143)

Steel Grade Identification Based on Improved Convolution Neural Network

REN Weijian1a,1b ,SONG Yue1a,1b ,CHNE Jianling2 ,REN Lu3 ,SUN Qinjiang4   

  1. 1a. School of Electrical Engineering & Information; 1b. Heilongjiang Provincial Key Laboratory of Networking & Intelligent Control,Northeast Petroleum University,Daqing 163318,China;2. Bohai Oil Research Institute,Tianjin Branch of CNOOC Company Limited,Tianjin 300452,China;3. Engineering Company,Offshore Oil Engineering Company Limited,Tianjin 300450,China;4. Tianjin Branch of CNOOC Company Limited,Tianjin 300452,China
  • Received:2017-05-06 Online:2018-05-24 Published:2018-07-25

摘要: 为获取样本的多样性特征,提出了一种改进的卷积神经网络结构。该网络中引入多层递归神经网络,利用卷积神经网络提取输入图像的浅层特征,同时利用卷积神经网络和递归神经网络并行提取高层特征,最后将两种网络学习到的特征进行融合输入到分类器中分类。利用迁移学习理论解决小样本集数据训练不足的问题,并将这种卷积神经网络结构应用于石油物资管线钢号识别中。实验探究了递归神经网络个数与卷积核个数对网络性能的影响,实验结果表明,改进的网络结构与其它网络进行对比,错误率降低了 3% 。

关键词: 递归神经网络, 小样本集, 卷积神经网络, 迁移学习

Abstract:  To extract diversity image features,an improved convolution neural network structure is proposed.The network introduced a multi-layer recursive neural network. Firstly,the network learned the shallow features of input images with convolution neural network,and then learned high-level features through the convolution neural network and multi-layer recursive neural network at the same time. Finally input two kinds of high-level features fusion into the classifier. This paper used the theory of transfer learning to solve the problem of the small sample training set data,and applied the improved convolution neural network structure to the oil pipeline steel grade identification. The experiments explore the influence of recursive neural network number and convolution kernels number separately,and make sure the final structure. The results show that the improved convolution neural network gets a low error rate of 3% .

Key words: convolution neural network, transfer learning, recursive neural network, small sample set

中图分类号: 

  • TP391. 4