基于SSVEP直接脑控机器人方向和速度研究

发布时间：2018-06-20 18:08

本文选题：脑控机器人 + 稳态视觉诱发电位　；参考：《自动化学报》2016年11期

【摘要】：直接用思维意图来控制机器人而没有大脑外周神经和肌肉的参与是人类的一个梦想,目前这一研究已成为国际前沿热点和突破点.传统的脑控机器人(Brain-controlled robot,BCR)主要控制其方向,而本文旨在探讨能够同时脑控机器人方向和速度的有效方法.采用可分类目标数多、单次识别率高且训练时间短的稳态视觉诱发电位(Steady state visual evoked potentials,SSVEP)脑机交互(Brain-computer/machine interaction,BCI/BMI)方法,为脑控机器人运动规划了向左、向右、前进和后退4个方向,设计了低速、中速和高速3级运动速度并组合了9个脑控指令;进而比较并优化了SSVEP刺激目标布局间距以及刺激目标闪烁时间,采用典型相关分析(Canonical correlation analysis,CCA)进行识别.结果表明恰当设置SSVEP刺激目标数及其布局间距和刺激目标闪烁时间,可以有效提高被试/用户直接脑控机器人的性能;优化的SSVEP刺激范式三结合适应SSVEP解码的典型相关分析,8名被试脑控机器人到达终点平均用时为2分40秒,最少用时1分29秒;同时,在脑控机器人运动过程中触碰障碍平均次数为0.88,最少碰触次数为0.本研究显示基于SSVEP的脑机交互可以作为直接脑控机器人灵活运动的一种可选方法,能够实现对机器人多个运动方向和多级速度的控制;也证实了适当增加刺激目标间距可以有效提高SSVEP-BCI脑控指令识别的正确率,说明了该脑控方法的性能与刺激被试的范式有关;再次验证了CCA算法在基于SSVEP的脑机交互中具有优良的效果.最后,为克服单一SSVEP范式存在的局限,本研究也尝试把该范式与运动想象相结合的混合范式用于脑控机器人方向和速度,并进行了初步的研究,表明可以进一步改善控制速度和提高被试舒适度.本文可望为基于SSVEP或与运动想象混合的脑机交互应用于分级或精细控制机器人方向和速度提供思路,并为直接脑控机器人技术推向实际应用打下一定的基础.
[Abstract]:It is a human dream to control robots directly with the intention of thinking without the participation of peripheral nerves and muscles of the brain. At present, this research has become a hot spot and breakthrough point in the international frontier. The traditional brain-controlled robot Brain-controlled robot (BCR) mainly controls its direction. This paper aims to explore an effective method to simultaneously control the direction and speed of the brain-controlled robot. The brain-computer interactive Brain-computer / computer interaction (BCI) method, which has a large number of classifiable targets, high single recognition rate and short training time, is used to plan the left, right, forward and backward directions for the movement of brain-controlled robot. The low speed, middle speed and high speed three-stage motion speed are designed and combined with 9 brain control instructions, and then the layout spacing and the flicker time of the stimulus target are compared and optimized, and the canonical correlation analysis method is used to identify them. The results show that the performance of directly brain-controlled robot can be effectively improved by setting the number of SSVEP stimulus targets and the spacing of the stimulus targets and the flicker time of the stimulus targets. Optimized SSVEP stimulation Paradigm 3 combined with canonical correlation analysis of adaptive SSVEP decoding, the average time for 8 brain-controlled robots to reach the end point was 2 minutes 40 seconds, and the minimum time was 1 minute 29 seconds. The average number of contact obstacles is 0.88 and the minimum contact number is 0. 8 in the movement of brain-controlled robot. This study shows that the brain-computer interaction based on SSVEP can be used as an alternative method for the direct brain-controlled robot to move flexibly, and it can realize the control of multi-direction and multi-stage velocity of the robot. It is also proved that the correct rate of SSVEP-BCI brain-controlled instruction recognition can be improved by increasing the target spacing, which indicates that the performance of the brain control method is related to the stimulator's paradigm. Finally, it is verified that CCA algorithm has good effect in brain computer interaction based on SSVEP. Finally, in order to overcome the limitations of the single SSVEP paradigm, this study also attempts to apply the hybrid paradigm combined with motion imagination to the direction and speed of the brain-controlled robot, and makes a preliminary study. The results showed that the control speed and comfort of the subjects could be further improved. This paper is expected to provide some ideas for the application of brain computer interaction based on SSVEP or hybrid with motion imagination in the direction and speed control of hierarchical or fine robot, and to lay a foundation for the practical application of direct brain-controlled robot technology.
【作者单位】：昆明理工大学信息工程与自动化学院;
【基金】：国家自然科学基金(61363043,61463024,81470084) 云南省应用基础研究计划(2013FB026) 云南省级人培项目(KKSY201303048) 云南省教育厅重点项目(2013Z130) 昆明理工大学脑信息处理与脑机交互融合控制(学科方向团队建设经费)资助~~
【分类号】：TP242

【相似文献】