语音交互技术在机械臂遥操作中的应用研究

Research on Application of Speech Interaction Technology in Teleoperation of Robotic Manipulator

目的研究语音交互在机械臂遥操作控制中的绩效与控制特性。方法基于机械臂显控界面工效学研究仿真平台、讯飞语音识别与合成SDK,研制了机械臂遥操作语音控制仿真实验测试平台。结合机械臂遥操作任务,针对性进行了语音识别率测试;开展了不同难度的舱段转位、大小臂组合及单关节位置任务的语音/触摸两种控制方式比较研究。结果对于机械臂半自主控制任务,在控制准确性相当的情况下,与触摸控制方式相比较,语音控制方式可以明显提高操作速度,降低控制难度,降低操作负荷、视觉负荷和认知加工负荷。结论对于较复杂、操作负荷较大的半自主控制任务,语音控制方式比触控方式具有更好的适用性,可明显提升控制绩效与体验,降低负荷。

Objective To study the human performance and control characteristics of speech interaction in the teleoperation tasks of robotic manipulator. Methods Based on the simulation platform for teleoperation ergonomic research and the i FLYTEK SDK of speech recognition and synthesis,a simulation test platform was developed for teleoperation of robotic manipulator in touch or speech interactive mode. According to the tasks of the robotic manipulator,the speech recognition rate was tested and the performances of three typical teleoperation tasks were compared for the two interactive methods. Results Compared with touch interaction,speech interaction could significantly increase operation speed,decrease operation difficulty,reduce visual load,operation load and recognition load under the same operation accuracy for the semi autonomous control tasks of robotic manipulator. Conclusion Speech interaction has better adaptability than touch interaction for complicated semi autonomous control tasks. It could optimize control performance,experience and work load.