摘要
数字电视发射机房处于高功率电磁辐射与多源噪声叠加环境,声学信号异常识别涉及射频干扰、电声耦合以及声源结构性扰动等复杂问题。针对这些问题,分析音频异常信号识别系统的总体架构,研究噪声抑制模块、特征提取模块、异常检测模块的设计方案,构建特征提取模型,并提出异常检测算法,旨在实现高信噪比、低延迟、广适配的结构化解耦识别路径。
Digital television transmitter room is in the environment where high-power electromagnetic radiation and multi-source noise are superimposed.Acoustic signal anomaly identification involves complex problems such as radio frequency interference,electro-acoustic coupling and structural disturbance of sound source.In order to solve these problems,the overall architecture of audio abnormal signal recognition system is analyzed,the design schemes of noise suppression module,feature extraction module and abnormality detection module are studied,the feature extraction model is constructed,and the abnormality detection algorithm is proposed,so as to realize the structural decoupling identification path with high signal-to-noise ratio,low delay and wide adaptability.
作者
梁传勇
LIANG Chuanyong(Yulin Branch of Guangxi Radio and Television Technology Center,Yulin 537000,China)
出处
《电声技术》
2025年第6期91-93,共3页
Audio Engineering
关键词
噪声抑制
异常检测
信号识别
noise suppression
anomaly detection
signal recognition
