基于语音识别的操作票智能防误研究  

Research on Intelligent Error Prevention of Operation Ticket Based on Speech Recognition

作　　者：陈汝昌 朱余启 朱润杭 莫熙 CHEN Ru-chang;ZHU Yu-qi;ZHU Run-hang;MO Xi(Yunnan Power Grid Co.,Ltd.,Kunming 650200 China)

机构地区：[1]云南电网有限责任公司,云南昆明650200

出　　处：《自动化技术与应用》2023年第9期68-71,83,共5页Techniques of Automation and Applications

摘　　要：深度卷积神经算法能够有效提高语音识别对于电力术语的识别率。通过深度学习算法训练语音识别的声学模型与语言模型,可以解决调度电力术语识别困难的情况,最终结合到与调度的相关应用中,有效地对操作票下令进行语音智能防误,帮助调度员减轻工作负担。Deep convolution neural algorithm can effectively improve the recognition rate of power terms.Through the deep learning algorithm to train the acoustic model and language model of speech recognition,it can solve the difficult situation of dispatching power term recognition.Finally,it can be combined with the relevant application of dispatching,effectively carry out voice intelligent error prevention for the operation order,and help the dispatcher reduce the workload.

关 键 词：操作票 智能防误 语音识别 调度电力 

分 类 号：TP391.42[自动化与计算机技术—计算机应用技术] TM272[自动化与计算机技术—计算机科学与技术]