A method for Collecting Measurement Data of Electric Energy Meters Based on Sound Prompt Technology
In order to solve the problems of low efficiency and poor real-time performance of traditional watt-hour meter data acquisition,a new method of watt-hour meter measurement data acquisition based on sound prompt technology is proposed.In this method,the speech broadcast signal of the electric energy meter is collected through a hypercentric directional microphone,and the Mel Frequency Cepstral Coefficient(MFCC)and Hidden Markov Model(HMM)are used for speech recognition and data decoding,and a Cyclic Redundancy Check(CRC)mechanism is introduced to ensure the data reliability.The experimental results show that the data acquisition accuracy of this method is 99.87%in the standard environment,and the acquisition time is 285 ms,and the accuracy can still be above 95%even in high noise environment,which has strong environmental adaptability and high practical value.
sound prompt technologyelectric energy meterdata collection
NETL
NSTL
万方数据
