基于表面肌电信号的人机交互软件设计

A Human-Computer Interaction Software Design Based on Surface EMG Signal

摘要:

设计了一种基于Qt的人机交互软件，用于从表面肌电信号中解码出手势，控制空间机械臂灵巧手作业。介绍了肌电信号解码手势并控制仿真灵巧手系统的组成，包括下位机肌电采集接口部分和上位机人机交互软件两部分。详细说明了人机交互软件的三个功能模块，即接收并显示16路神经接口向上位机发送的肌电信号、对肌电信号进行实时手势解码以及控制仿真灵巧手。分析了软件设计过程中的几个关键技术，信号与槽机制、多线程与多进程结合、UDP通信等。最后，设计了基于肌电信号解码三种手势并控制仿真灵巧手的实时实验，手势识别率在98%以上，控制延迟为200ms左右。实验结果表明，人机交互软件运行稳定，功能齐全，在航天遥操作人机交互系统中具有应用前景。

Abstract:

A Qt-based human-computer interaction software is designed to decode gestures from surface EMG signals and control the dexterous hand operation of the space manipulator. The composition of the EMG signal decoding gesture and control simulation dexterous hand system is introduced, including the EMG collection interface part of the lower computer and the human-computer interaction software of the upper computer. The three functional modules of human-computer interaction software are explained in detail, namely receiving and displaying the EMG signal sent by the 16-way neural interface to the upper computer, real-time gesture decoding of the EMG signal, and control of simulated dexterous hands. Several key technologies in the software design process are analyzed, such as the signal and slot mechanism, the combination of multithreading and multiprocessing, and UDP communication. Finally, a real-time experiment was designed to decode three gestures based on the EMG signal and control the simulated dexterous hand. The gesture recognition rate was over 98%, and the control delay was about 200ms. Experimental results show that the human-computer interaction software runs stably and has complete functions, and has application prospects in human-computer interaction systems of aerospace teleoperation.