|本期目录/Table of Contents|

[1]方维岚,陆正卿,曹 鑫,等.基于声纹识别的蒸汽泄漏报警监控系统[J].工业仪表与自动化装置,2024,(04):32-35+51.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2024.04.007]
　FANG Weilan,LU Zhengqing,CAO Xin,et al.A system for monitoring steam leakage alarm based on voiceprint recognition[J].Industrial Instrumentation & Automation,2024,(04):32-35+51.[doi:DOI:10.19950/j.cnki.cn61-1121/th.2024.04.007]
点击复制

基于声纹识别的蒸汽泄漏报警监控系统(PDF)

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2024年04期
页码:
32-35+51
栏目:
出版日期:
2024-08-15

文章信息/Info

Title:
A system for monitoring steam leakage alarm based on voiceprint recognition
文章编号:
1000-0682（2024）04-0032-04
作者:
方维岚; 陆正卿; 曹 鑫; 等
（上海烟草集团有限责任公司，上海 200082）
Author(s):
FANG Weilan;  LU Zhengqing;  CAO Xin;  et al
(Shanghai Tobacco Group Co., Ltd., Shanghai 200082, China)
关键词:
蒸汽泄漏; 声源定位; 声纹识别; Mel倒谱系数; 长短时记忆神经网络
Keywords:
steam leakage;  sound source localization;  voiceprint recognition;  MFCC;  LSTM
分类号:
TB52+6
DOI:
DOI:10.19950/j.cnki.cn61-1121/th.2024.04.007
文献标志码:
A
摘要:
蒸汽管道在工业企业中随处可见，并且是相当重要的设备之一。一旦发生蒸汽泄漏，轻则浪费能源，影响生产质量，重则造成安全事故。由于蒸汽管道往往位于工厂隐蔽的位置，一旦泄漏，除了日常巡检，没人会发现。为了及时发现蒸汽泄漏，研制了1套基于声纹识别的蒸汽泄漏报警监控系统。由于蒸汽管道处于密闭空间，且周围没有其他设备运行，几乎不存在声音干扰问题。通过整列麦克风，利用声源定位算法对气体泄漏频段内的声音进行识别和定位。实验结果表明，通过特定频段和声纹识别，能够有效提高蒸汽泄漏的识别率，实现蒸汽泄漏的自动识别报警。
Abstract:
Steam pipelines are ubiquitous in industrial enterprises and are one of the most important equipment. Once a steam leak occurs, it can waste energy, affect production quality, and even cause safety accidents. Due to the fact that steam pipelines are often located in hidden locations in factories, once a leak occurs, no one will notice it except for daily inspections. So, in order to detect steam leaks in a timely manner, a steam leak alarm and monitoring system based on voiceprint recognition has been developed. Due to the steam pipeline being in a closed space and no other equipment operating around it, there is almost no problem of sound interference. Identify and locate the sound within the gas leakage frequency band using a sound source localization algorithm through a whole array of microphones. The experimental results show that through specific frequency bands and voiceprint recognition, the recognition rate of steam leaks can be effectively improved, and automatic recognition and alarm of steam leaks can be achieved.

参考文献/References:

[1] 周萍,沈昊,郑凯鹏.基于MFCC与GFCC混合特征参数的说话人识别[J].应用科学学报,2019,37(1): 24-32.

[2] Kasiprasad M, Panyam N S, Maloji S. MFCC-GMM based accent recognition system for Telugu speech signals[J]. International Journal of Speech Technology, 2016(19): 87-93.
[3] Ankur M, Divya K, R.K.Agarwal. Speaker Recognition for Hindi Speech Signal using MFCC-GMM Approach[C]. 6th International Conference on Smart Computing and Communications (ICSCC), 2017, 880-887.
[4] GREFF K, SRIVASTAVA R K, KOUTNIK J, et al. LSTM: A search space odyssey[J]. IEEE Transactions on Neural Networks ＆ Learning Systems,2017,29(10): 2222-2232.
[5] Himadri M, Sk Md Obaidullah, K.C.Santosh, et al. A lazy learning-based language identification from speech using MFCC-2 features[J]. International Journal of Machine Learning and Cybernetics,2020(11): 1-14.
[6] Sakshi G, Ravi S S Rajesh K. S, et al. Deep Learning Bidirectional LSTM based Detection of Prolongation and Repetition in Stuttered Speech using Weighted MFCC[J]. International Journal of Advanced Computer Science and Applications,2020,11(9):345-456.
[7] Samia A El-Moneim, M. A. Nassar, et al. Text-independent speaker recognition using LSTM-RNN and speech enhancement[J]. Multimedia Tools and Applications, 2020(79): 24013– 24028.
[8] 王鑫,吴际,刘超,等.基于LSTM循环神经网络的故障时间序列预测[J].北京航空航天大学学报,2018,44(4): 772-784.
[9] 龚安,马光明,郭文婷,等.基于LSTM循环神经网络的核电设备状态预测[J].计算机技术与发展,2019,29(10):41-45.
[10] 李俊卿,陈雅婷.基于LSTM网络的同步电机励磁绕组匝间短路故障预警[J].电力科学与工程,2020,36(6):37-42.
[11] 牟俊杰,姚刚,孙涛.基于CNN-LSTM神经网络的声纹识别系统设计[J].电子技术应用,2021,47(3): 75-78.
[12]王志成,张玉一,巴天星,等.面向CIM和动态交通分析的多源异构数据融合技术研究[J].电子设计工程,2024,32(8):68-72.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期：2024-03-04第一作者：方维岚（1969—），男，上海人，工业电气自动化专业硕士，高级工程师，主要研究方向为工业自动控制与建筑智能化等。E-mail：dengliao79855995@163.com
更新日期/Last Update: 1900-01-01