声纹识别下的电网调度指令自动化校验方法  

Automatic verification method of power grid dispatching instruction based on voiceprint recognition

作　　者：梁铃 马建伟 周忠强 王平 陈智祺 LIANG Ling;MA Jianwei;ZHOU Zhongqiang;WANG Ping;CHEN Zhiqi(Power Dispatching Control Center of Guizhou Power Grid Company,Guiyang 550000,China)

机构地区：[1]贵州电网公司电力调度控制中心,贵州贵阳550000

出　　处：《电子设计工程》2024年第18期136-140,共5页Electronic Design Engineering

基　　金：贵州电网有限责任公司科技项目(060400KK42160025)。

摘　　要：在校验电网调度指令的过程中,为实现对噪声信号的准确识别,使电力网络保持较为积极的运行状态,提出声纹识别下的电网调度指令自动化校验方法。通过声纹信号筛选结果确定特征参量取值范围,并求解目标源函数,实现对电网调度指令的识别。分别计算降频系数与帧长参量,推导加权降维条件,完成声纹识别下电网调度指令自动化校验方法的设计。实验结果表明,所提算法的噪声信号电压和电流检测结果与真实值的最大差值分别为3.5 V和2.0 A,符合准确识别噪声信号的电网调度指令校验需求。In the process of verifying power grid dispatching instructions,in order to realize accurate recognition of noise signals and keep the power network in a more active operation state,an automatic verification method of power grid dispatching instructions based on voiceprint recognition is proposed.Through the voiceprint signal screening results,the range of characteristic parameters is determined,and the target source function is solved to realize the recognition of power grid dispatching instructions.The frequency reduction coefficient and frame length parameters are calculated respectively,and the weighted dimensionality reduction conditions are derived to complete the design of automatic verification method for power grid dispatching instructions under voiceprint recognition.The experimental results show that the maximum differences between the noise signal voltage and current detection results of the proposed algorithm and the true values are 3.5 V and 2.0 A,respectively,which meets the requirements of power grid dispatching command verification for accurately identifying noise signals.

关 键 词：声纹识别 电网调度指令 自动化校验 目标源函数 降频系数 加权降维条件 

分 类 号：TN912[电子电信—通信与信息系统]