Human-Machine Interaction (HMI) and Brain-Computer Interface (BCI) are evolving technologies that show the great potentials to extract and utilize humans’ intents in controlling smart machines. However, existing HMI and BCI technologies are limited in terms of (1) the number of Degrees- of-Freedom (DoF) to be controlled and (2) the ways the performance of BCI-enabled control systems are verified and validated. This study aimed to explore the solutions to addree both of above concerns; we proposed a hybrid control system that is capable of training, detecting, and interpreting humans’ intents, and utilizing humans’ intents in real-time controls of smart machines. More specifically, the system acquired brain signals in the form of Electroencephalography (EEG) by an Emotiv Epoc X and processed these signals to detect and extract humans’ intents in real-time machine controls. To cope with the frequency difference of humans’ thinking and machine motion controls, we developed a hybrid control module to fuse humans’ and machine's intelligence so that low-frequency humans’ intents could be used in real-time machine controls. The system was prototyped and verified experimentally. The system was verified to achieve the accuracy of over 90 % in recognizing humans’ intents and controlling a robot by the operator's intents with a satisfactory responding time and accuracy.

中文翻译：

---

智能机器的人机协作控制——提出的混合控制


人机交互 (HMI) 和脑机接口 (BCI) 是不断发展的技术，显示出提取和利用人类意图来控制智能机器的巨大潜力。然而，现有的 HMI 和 BCI 技术在以下方面受到限制：(1) 要控制的自由度 (DoF) 数量，以及 (2) 验证和验证支持 BCI 的控制系统的性能的方式。本研究旨在探索解决上述两个问题的方案；我们提出了一种混合控制系统，能够训练、检测和解释人类意图，并利用人类意图实时控制智能机器。更具体地说，该系统通过 Emotiv Epoc X 以脑电图 (EEG) 形式获取大脑信号，并处理这些信号以检测和提取实时机器控制中的人类意图。为了应对人类思维和机器运动控制的频率差异，我们开发了一种混合控制模块来融合人类和机器的智能，使低频人类意图可以用于实时机器控制。该系统进行了原型设计并进行了实验验证。经验证，系统在识别人类意图、通过操作者意图控制机器人方面的准确率达到90%以上，并且响应时间和准确度令人满意。