一种基于EEMD的异常声音识别方法

doi:10.13190/j.jbupt.2020-075

北京邮电大学学报 ›› 2020, Vol. 43 ›› Issue (5): 112-117.doi: 10.13190/j.jbupt.2020-075

一种基于EEMD的异常声音识别方法

韦娟¹, 顾兴权¹, 宁方立²

1. 西安电子科技大学通信工程学院, 西安 710071;
2. 西北工业大学机电学院, 西安 710072

收稿日期:2020-06-25 发布日期:2021-03-11
作者简介:韦娟(1973-),女,副教授,E-mail:weijuan@xidian.edu.cn.
基金资助:
国家自然科学基金项目（51675425）；陕西省重点研发计划项目（2018GY-181，2020ZDLGY06-09）

An Abnormal Sound Recognition Method Based on EEMD

WEI Juan¹, GU Xing-quan¹, NING Fang-li²

1. School of Communication Engineering, Xidian University, Xi'an 710071, China;
2. School of Mechanical Engineering, Northwestern Polytechnical University, Xi'an 710072, China

Received:2020-06-25 Published:2021-03-11

摘要/Abstract

摘要： 为了优化组合特征在异常声音识别中的效率，提出一种用集合经验模态分解（EEMD）对异常声音帧信号进行有效性检测和提取多层特征的算法.首先对异常声音帧信号进行集合经验模态分解，得到固有模态函数；然后根据给定的固有模态函数层数阈值，对该帧信号进行有效性检测；再对有效帧信号的每一层固有模态函数提取梅尔频率倒谱系数、翻转梅尔频率倒谱系数、线性预测倒谱系数、短时能量和能量比，并将它们归一化后拼接成多层特征.根据提取的特征，用深度卷积神经网络实现异常声音识别分类.仿真结果表明，提出的新方法在4类异常声音识别中的识别率可以达到98.65%.

关键词: 异常声音识别, 集合经验模态分解, 多层特征, 深度卷积神经网络

Abstract: In order to optimize the efficiency of combined features in abnormal sound recognition,an algorithm for detecting the effectiveness of abnormal sound frame signals and extracting multi-layer features using ensemble empirical mode decomposition (EEMD) is proposed. Firstly,an ensemble empirical mode decomposition is performed on the abnormal sound frame signal to obtain the intrinsic model function,and then the validity of the frame signal is tested according to the given layer threshold of the intrinsic modal function. Finally,the Mel frequency cepstral coefficients,the inverted Mel frequency cepstral coefficients,the linear prediction cepstral coefficients,the short-time energy and energy ratio are extracted for each layer of the intrinsic modal function of the effective frame signal,and then all of them are normalized and spliced into multi-layer feature. According to the extracted features,the deep convolutional neural network is used to realize the classification and recognition of abnormal sound. Simulations show that the proposed new method can achieve a recognition rate of 98.65% in four types of abnormal sound recognition.

Key words: abnormal sound recognition, ensemble empirical mode decomposition, multi-layer feature, deep convolutional neural network

中图分类号:

TP391

韦娟, 顾兴权, 宁方立. 一种基于EEMD的异常声音识别方法[J]. 北京邮电大学学报, 2020, 43(5): 112-117.

WEI Juan, GU Xing-quan, NING Fang-li. An Abnormal Sound Recognition Method Based on EEMD[J]. Journal of Beijing University of Posts and Telecommunications, 2020, 43(5): 112-117.

参考文献

[1] 陈志全, 杨骏, 乔树山. 基于EEMD的异常声音特征提取[J]. 计算机与数字工程, 2016, 44(10):1875-1879. Chen Zhiquan, Yang Jun, Qiao Shushan. Abnormal sound feature extraction based on EEMD[J]. Computer and Digital Engineering, 2016, 44(10):1875-1879.
[2] Xu Jining, Yao Xiaoxin. Abnormal sound recognition with audio feature combination and modified GMM[C]//Proceedings of the 32nd Chinese Control Conference. Xi'an:IEEE Press, 2013:4582-4585.
[3] Pedroza Ramirez A D, De La Rosa Vargas J I, Valdez R R, et al. A comparative between Mel frequency cepstral coefficients and inverse Mel frequency cepstral coefficients features for an automatic bird species recognition system[C]//2018 IEEE Latin American Conference on Computational Intelligence. Gudalajara:IEEE Press, 2018.
[4] 李颀, 白雨尼, 王丹聪. 基于小波包分析的玻璃破碎声音识别系统设计[J]. 计算机测量与控制, 2018, 26(1):168-172. Li Qi, Bai Yuni, Wang Dancong. Design of class breaking sound recognition system based on wavelet packet transform[J]. Computer Measurement and Control, 2018, 26(1):168-172.
[5] Yan Guolin, Wang Mei, Liu X, et al. Sound event recognition based in feature combination with low SNR[C]//2019 International Conference on Artificial Intelligence and Advanced Manufcturing. Dublin:IEEE Press, 2019:109-114.
[6] 韦娟, 张芃楠, 岳凤丽, 等. 基于PSO-PF算法的SVM识别方法及其在异常声音中的应用[J]. 北京邮电大学学报, 2019, 42(3):58-63. Wei Juan, Zhang Pengnan, Yue Fengli, et al. Recognition and application of abnormal sound via SVM based on PSO-PF[J]. Journal of Beijing University of Posts and Telecommunications, 2019, 42(3):58-63.
[7] Zhang Shangyue, Liu Yuanyuan, Yang Gongliu. EMD interval thresholding denoising based on correlation coefficient to select relevant modes[C]//2015 34th Chinese Control Conference. Hangzhou:IEEE Press, 2015:4801-4806.
[8] 杨恭勇, 周小龙, 李家飞, 等. 局部Hilbert边际能量谱在滚动轴承故障诊断中的应用[J]. 东北电力大学学报, 2017, 37(2):77-81. Yang Gongyong, Zhou Xiaolong, Li Jiafei, et al. A study of rolling bearing fault diagnosis based on local Hilbert marginal energy spectrum[J]. Journal of Northeast Electric Power University, 2017, 37(2):77-81.
[9] Winursito A, Hidayat R, Beji A, et al. Improvement of MFCC feature extraction accuracy using PCA in indonesian speech recognition[C]//2018 International Conference on Information and Communications Technology. Yogyakarta:IEEE Press, 2018:379-383.
[10] Chakroborty S, Roy A, Majumdar S, et al. Capturing complementary information via reversed filter bank and parallel implementation with MFCC for improved text-independent speaker identification[C]//2007 International Conference on Computing:Theory and Applications. Kolkata:IEEE Press, 2007.
[11] Gupta H, Gupta D. LPC and LPCC method of feature extraction in speech recognition system[C]//2016 6th International Conference Cloud System and Big Data Engineering. Noida:IEEE Press, 2016:498-502.
[12] Eltiraifi O, Elbasheer E, Nawari M. A comparative study of MFCC and LPCC features for speech activity detection using deep belief network[C]//2018 International Conference on Computer Control, Electrical and Electronics Engineering. Khartoum:IEEE Press, 2018.
[13] Liu Gang, He Wei, Jin Bicheng. Feature fusion of speech emotion recognition based on deep learning[C]//2018 International Conference on Network Infrastructure and Digital Content. Guiyang:IEEE Press, 2018:193-197.
[14] Nikhitha M, Roopa Sir S, Uma Maheswari B, et al. Fruit recognition and grade of disease detection using inception V3 model[C]//2019 3rd International Conference on Electronics, Communication and Aerospace Technology. Coimbatore:IEEE Press, 2019:1040-1043.

一种基于EEMD的异常声音识别方法

An Abnormal Sound Recognition Method Based on EEMD

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 5

编辑推荐

Metrics

本文评价

[1]	许同乐, 孟良, 孔晓佳, 苏元浩, 孙砚飞. 基于EEMD的ICNN故障诊断方法[J]. 北京邮电大学学报, 2022, 45(2): 110-116.
[2]	李喆, 李建增, 王哲. 基于先验优化的一致性模糊盲复原算法[J]. 北京邮电大学学报, 2019, 42(2): 63-69.
[3]	邹国锋, 傅桂霞, 王科俊, 高明亮, 申晋. 自适应深度卷积神经网络模型构建方法[J]. 北京邮电大学学报, 2017, 40(4): 98-103.
[4]	李亚, 王广润, 王青. 基于深度卷积神经网络的跨年龄人脸识别[J]. 北京邮电大学学报, 2017, 40(1): 84-88,110.
[5]	李玲慧, 杜军平, 梁美玉, 任楠, JangMyung Lee. 基于时空特征和神经网络的视频超分辨率算法[J]. 北京邮电大学学报, 2016, 39(4): 1-6.