A Rub-Impact Recognition Method Based on Improved Convolutional Neural Network

作     者：Weibo Yang Jing Li Wei Peng Aidong Deng 

作者机构：College of Civil AviationNanjing University of Aeronautics and AstronauticsNanjing210016China School of Information EngineeringNanjing Audit UniversityNanjing211815China National Engineering Research Center of Turbo-generator VibrationSoutheast UniversityNanjing210009China 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2020年第63卷第4期

页      面：283-299页

核心收录：

学科分类：0808[工学-电气工程] 0809[工学-电子科学与技术（可授工学、理学学位）] 07[理学] 08[工学] 071102[理学-系统分析与集成] 0831[工学-生物医学工程（可授工学、理学、医学学位）] 0711[理学-系统科学] 0805[工学-材料科学与工程（可授工学、理学学位）] 081101[工学-控制理论与控制工程] 0701[理学-数学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 0801[工学-力学（可授工学、理学学位）] 081103[工学-系统工程] 

基　　金：The authors would like to acknowledge the Six Talent Peaks Project in Jiangsu Province[XCL-CXTD-007] China Postdoctoral Science Foundation[2018M630559]for their financial support in this project 

主　　题：Acoustic emission signal deep learning convolutional neural network spectral features rub-impact 

摘      要：Based on the theory of modal acoustic emission(AE),when the convolutional neural network(CNN)is used to identify rotor rub-impact faults,the training data has a small sample size,and the AE sound segment belongs to a single channel signal with less pixel-level information and strong local *** to the convolutional pooling operations of CNN,coarse-grained and edge information are lost,and the top-level information dimension in CNN network is low,which can easily lead to *** solve the above problems,we first propose the use of sound spectrograms and their differential features to construct multi-channel image input features suitable for CNN and fully exploit the intrinsic characteristics of the sound ***,the traditional CNN network structure is improved,and the outputs of all convolutional layers are connected as one layer constitutes a fused feature that contains information at each layer,and is input into the network’s fully connected layer for classification and *** indicate that the improved CNN recognition algorithm has significantly improved recognition rate compared with CNN and dynamical neural network(DNN)algorithms.