摘要
针对近年来机器人模仿类似人类笔迹的纸笔书写面临的反应速度慢、精度低及交互繁琐的问题,提出并实现了由4个舵机驱动的四自由度机械臂关节结构,设计了根据书写笔画通过自学习完成的包括舵机编号、角度、动作运行时间等信息构成的动作代码,实现了机械臂书写的毫米级稳定控制。在此基础上,设计实现了一套基于语音控制的智能书写机器人系统,即通过语音控制小型机械臂完成文字书写。该系统采用树莓派3B为主控与32路伺服舵机控制板相结合构成控制平台,实现了语音识别、发音比对纠错、语音文本转换、即时语音书写、书写字体切换、书写精度控制等模块和功能,系统中精密细小部件利用3DSMAX绘图3D打印制成。系统测试结果表明,所设计系统的语音平均识别反馈文本时间为2.36 s,单字平均书写时间为2.12 s,与一般情况下人手书写速度持平。
Concerning the problems of slow response,low accuracy,and tedious interaction when robots mimick paper writing similar to human handwriting in recent years,a four-degree-of-freedom robotic arm joint structure driven by 4 servos was proposed and a code including information such as the servo number,angle and motion running time was realized which can control the robot arm in millimeter-level stability.An intelligent writing robot system with speech control was designed and implemented,and a small robotic arm was designed to complete the writing.A Raspberry Pi 3B and a servo control board were combined to form a control module,which realized voice recognition,pronunciation comparison and correction,voice text conversion,instant voice writing,writing font switching,writing accuracy control,etc.Precision small parts in the system were made by 3D printing with 3 DSMAX drawing.The system test results show that,the average speech recognition time of the designed system is 2.36 s and the average writing time of a single character is 2.12 s,which is comparative with handwriting under normal circumstances.
作者
张玉琢
郑世珏
潘磊
ZHANG Yuzuo;ZHENG Shijue;PAN Lei(College of Computer Science and Technology,Central China Normal University,Wuhan Hubei 430079,China)
出处
《计算机应用》
CSCD
北大核心
2020年第S01期284-288,共5页
journal of Computer Applications
基金
国家社会科学基金资助项目(17BYY111)
大学生创新创业训练计划项目(105112019010019)。
关键词
书写机器人
语音识别
语音控制
四自由度机械臂
伺服舵机
writing robot system
speech recognition
speech control
four degree of freedom manipulator
servo actuator