摘要
针对城市轨道交通模式,为提高地铁服务型机器人语音控制效果,设计一款融入语义理解技术的服务型机器人语音控制系统。系统硬件部分由利用附带模数转换器的音频输入输出电路、以S5P4418为主芯片的存储器电路、通用异步收发传输器串口与无线Wi-Fi的通信电路等组件组成;软件部分为采用连续语音基元识别器识别语音文本,根据似然得分排序,明确关键词,结合控制语料结构实现语义理解,基于socket通信节点与机器人操作系统节点的发布主题,经node节点与voice_nav节点完成语音控制。实验结果表明,该系统能够给出有效反馈,且精准度较高,可行性较为理想。
In order to improve the speech control effect of subway service robot in the urban rail transit mode,a speech control system of service robot with semantic understanding technology is designed.The hardware part of the system consists of audio input and output circuit with analog-to-digital converter,memory circuit with main chip S5P4418,communication circuit of universal asynchronous receiver-transmitter serial port and wireless Wi-Fi,etc.The software part adopts the speech text that is recognized by continuous speech primitive recognizer,sorted according to likelihood score to define keywords.Semantic understanding is realized by combining with control corpus structure.Based on the publishing theme of socket communication node and robot operating system node,speech control is completed through node node and voice_nav node.Experiment results show that the system can give effective feedback with high accuracy and ideal feasibility.
作者
梁立
刘宁
徐炜
应擎天
LIANG Li;LIU Ning;XU Wei;YING Qing-tian(Operating Branch,Ningbo Metro Group Co.,Ltd.,Ningbo 315000,Zhejiang Province,China)
出处
《信息技术》
2022年第11期130-135,共6页
Information Technology
关键词
语义理解
机器人
语音控制
关键词提取
semantic understanding
robot
voice control
keywords extraction