Human operator control of brain-actuated robot steering based on electroencephalograph (EEG)-signals is a complex behavior consisting of surroundings perceiving, decision making, and commands issuing and differs among individual operators. However, no existing models allow decoupling the user from the loop to improve the system design and testing process, which can capture such behavior of a brain-actuated robot. To address this problem, in this paper, we propose an operator brain-controlled steering model consisting of an operator decision model based on the queuing network (QN) cognitive architecture and a brain-machine interface (BMI) performance model. The QN-based operator decision model can mimic the human decision process with the individual operator differences considered. The new BMI performance model is built to represent the varied accuracy of BMI during brain-controlled direction operations. Furthermore, the model is simulated and validated against the results of human operator-in-the-loop experiments. The results show that the proposed model can reproduce the behavior of human operators thanks to its similar direction control performance.

中文翻译：

---

大脑驱动的移动机器人转向的人类操作员行为建模

基于脑电图（EEG）信号的人类操作员对脑部驱动的机器人转向的控制是一种复杂的行为，包括周围环境的感知，决策和命令的发布，并且各个操作员之间存在差异。但是，没有现有的模型允许将用户从循环中解耦出来以改善系统设计和测试过程，从而可以捕获大脑驱动的机器人的这种行为。为了解决这个问题，在本文中，我们提出了一种操作员大脑控制的转向模型，该模型由基于排队网络（QN）认知体系结构的操作员决策模型和脑机接口（BMI）性能模型组成。基于QN的操作员决策模型可以模仿考虑了各个操作员差异的人工决策过程。建立了新的BMI性能模型，以表示在大脑控制的方向操作过程中BMI的准确性。此外，该模型是根据人工在环实验的结果进行仿真和验证的。结果表明，该模型具有相似的方向控制性能，可以重现操作员的行为。