摘要
针对输电铁塔螺栓松动问题,利用声纹识别原理与方法,对铁塔部分螺栓在紧固松动状态下的声音信号差异特征进行分析对比,建立了基于锤击信号MFCC特征向量提取与判定的声纹识别模型,主要包括声纹训练、声纹识别、结果输出三部分模块,搭建了声纹识别系统,通过试验基地铁塔螺栓人工松动与紧固状态试验识别,验证了该系统在识别铁塔螺栓松动与紧固状态的可靠性。
Aiming at the looseness of transmission tower bolts,this paper uses using the principle and method of voiceprint recognition to analyze and compare the difference characteristics of sound signals of some tower bolts in the state of fastening and looseness.A voiceprint recognition model based on MFCC feature vector extraction and judgment of hammering signal is constructed,which mainly includes three modules i.e.,voiceprint training module,voiceprint recognition module and result output module.A voiceprint recognition system was built,and the reliability of the system in identifying the looseness and fastening state of tower bolts was verified through the artificial looseness and fastening state test identification of the tower bolts in the test base.
作者
李文彬
冯砚厅
王勇
李晓康
吕亚东
LI Wenbin;FENG Yanting;Wang Yong;Li Xiaokang;LV Yadong(State Grid Hebei Energy Technology Service Co.,Ltd.,Shijiazhuang 050021,China;State Grid Hebei Electric Power Research Institute,Shijiazhuang 050021,China)
出处
《河北电力技术》
2022年第6期63-67,共5页
Hebei Electric Power
基金
国网河北能源技术服务有限公司科技项目(TSS2021-07)。
关键词
输电铁塔
螺栓松动
声纹识别
transmission tower
bolt looseness
voiceprint recognition
