Application Research Of Voiceprint Recognition Technology In Cable Ditch External Force Damage Early Warning System

At present,with the acceleration of the urbanization and the rapid economic development,urban electricity consumption has increased dramatically.In order to protect the image of the city and save land resources,reduce accidents that could threaten residents' safety,the distribution network overhead has been gradually replaced by underground cables.Therefore,it is a severe challenge to detect the safety of underground cable trenches.It needs to design an automated intelligent cable trench external force damage early warning system.With the upgrading of computer software and hardware and the rapid development of machine deep learning,big data and the Internet of Things,especially the emergence of new feature extraction algorithms,new models of speech and acoustics,and the widespread application of retrieval and adaptive algorithms,voiceprint recognition technology Based on this background,it is proposed to the research and application of voiceprint recognition technology in cable trench external force damage early warning system.Firstly,in order to solve the problem of preventing cable trench breakage,through the research on voiceprint recognition technology,this paper proposes the research topic: the application research of voiceprint recognition technology in cable trench external force damage early warning system.Then it introduced the software and hardware architecture of the cable trench external force damage early warning system.The vibration signals under different conditions are studied in depth.According to the characteristics of the vibration signals,it analyzed the feasibility of identifying external force damage vibration signals based on voiceprint recognition technology.Then,the vibration signal of external force destruction was collected experimentally,and the data was preprocessed,including denoising,pre-emphasis,windowing framing,and endpoint detection.According to the signal characteristics,the MFCC characteristic parameters of the signal and its first-order second-order difference parameters Weighted average with pitch period to obtain fusion parameters.Gaussian mixture model(GMM-UBM)the most commonly used model in voiceprint recognition technology,is used for pattern recognition.Firstly preprocess the external force damage signal,extract the pitch period parameters,MFCC characteristic parameters and their first-order second-order difference parameters,train a general background model of the vibration signal,use the target vibration signal to train a Gaussian mixture model,and update the general background model to get the final Gaussian mixture-universal background model.Input the signal to be identified and perform pattern matching to obtain the recognition result,and realize the function of the external force to destroy the early warning system.Finally,it conducted field test of the external force damage early warning system that based on voiceprint recognition technology.By testing the real-time nature of the early warning system,alarm level,positioning accuracy and other requirements,it was demonstrated that the recognition mode based on voiceprint recognition technology can be applied to external force of cable trench Damage early warning system.