基于声纹识别技术的电力调度系统设计  

Design of Power Dispatching System Based on Voiceprint Recognition Technology

作　　者：沈亚玲 SHEN Yaling(State Grid Jiangsu Electric Power Co.,Ltd.,Yancheng Dafeng District Power Supply Branch,Yancheng 224100,China)

机构地区：[1]国网江苏省电力有限公司盐城市大丰区供电分公司,江苏盐城224100

出　　处：《电声技术》2024年第12期1-3,共3页Audio Engineering

摘　　要：为提升电力调度的效率与安全性,探讨声纹识别技术在电力调度系统中的应用。采用梅尔频率倒谱系数(Mel Frequency Ceptral Coefficient,MFCC)特征提取与高斯混合模型(Gaussian Mixture Model,GMM)建模方法,结合大词汇量连续语音识别(Large Vocabulary Continuous Speech Recognition,LVCSR)技术和改进的蚁群优化(Ant Colony Optimization,ACO)算法,设计基于声纹识别的电力调度系统。实验结果显示,该系统在身份验证和指令处理方面表现出色,具备较高的准确率与响应速度,能够有效保障电网运行的安全性与稳定性。In order to improve the efficiency and safety of power dispatching,the application of voiceprint recognition technology in power dispatching system is discussed.Mel Frequency Cepstral Coefficient(MFCC) feature extraction and Gaussian Mixture Model(GMM) modeling method are adopted,Combining Large Vocabulary Continuous Speech Recognition(LVCSR) technology and improved Ant Colony Optimization(ACO) algorithm,a power dispatching system based on voiceprint recognition is designed.The experimental results show that the system performs well in identity verification and instruction processing,and has high accuracy and response speed,which can effectively guarantee the security and stability of power grid operation.

关 键 词：声纹识别 电力调度 梅尔频率倒谱系数(MFCC) 

分 类 号：TM73[电气工程—电力系统及自动化]