摘要
受火场环境中极端噪声的影响,消防应急装备接收到的语音信号信噪比较低,严重制约了语音识别与交互控制的准确性与可靠性。为此,提出一种窄带物联网(Narrow Band Internet of Things,NB-IoT)环境下的消防应急装备抗噪语音交互控制方法。研究首先从系统组成结构和通信协议两方面构建适用于消防场景的NB-IoT环境,依托该环境,将语音信号传输至消防应急装备控制终端。在信号处理阶段,利用子载波级加权抑制算法完成对信号的初步降噪;随后,通过固定增益的方式补偿自由空间路径损耗,达到语音交互信号深度抗噪的目的。最后,通过平滑滤波、特征提取与匹配处理,识别抗噪语音交互信号的内容。根据识别结果驱动控制程序,完成抗噪语音交互控制。实验结果表明:应用该方法后,语音信号的传输速率在121 Mb/s以上,满足设定要求。与传统方法相比,应用该方法信号的信噪比显著提升,可保持在66~82 dB之间,交互控制成功率更接近1.0,说明该方法具有更优的抗噪语音交互控制性能。
Due to the extreme noise in the fire scene,the signal-to-noise ratio of the received voice signals by fire emergency equipment is low,which seriously restricts the accuracy and reliability of voice recognition and interactive control.Therefore,this paper proposes a noise-resistant voice interaction control method for fire emergency equipment in a narrow band Internet of Things(NB-IoT)environment.The study first constructs an NB-IoT environment suitable for fire scenarios from the aspects of system composition and communication protocol.Based on this environment,the voice signal is transmitted to the control terminal of the fire emergency equipment.In the signal processing stage,a subcarrier-level weighted suppression algorithm is used to complete the initial noise reduction of the signal.Subsequently,free space path loss is compensated by a fixed gain method to achieve deep noise resistance of the voice interaction signal.Finally,the content of the noise-resistant voice interaction signal is identified through smoothing filtering,feature extraction,and matching processing.The control program is driven by the recognition result to complete the noise-resistant voice interaction control.Experimental results show that after applying this method,the transmission rate of the voice signal is above 121 Mb/s,meeting the set requirements.Compared with traditional methods,the signal-to-noise ratio of the applied method is significantly improved,remaining between 66 and 82 dB,and the interactive control success rate is closer to 1.0,indicating that this method has superior noise-resistant voice interactive control performance.
作者
丁晨
DING Chen(Emergency Response and Rescue Center of Petro China Xinjiang,Karamay 834000,China)
出处
《国外电子测量技术》
2025年第11期301-307,共7页
Foreign Electronic Measurement Technology
关键词
窄带物联网
消防应急装备
语音交互
抗噪语音
交互控制
narrow band Internet of Things
fire emergency equipment
voice interaction
noise-resistant voice
interactive control
