基于SSVEP的脑控小车系统的研究

Brain-controlled car system based on SSVEP

文中设计了一种基于SSVEP的脑控小车分级速度和方向控制的系统,存在向左、向右、向后、向前(一级、二级、三级)6个脑控命令。在刺激范式、刺激时间、空间布局三方面进行了优化,实验表明扩大刺激目标间距、加强刺激时间能够提高目标识别准确率。在脑电解码方面,采用HHT(Hibert-Huang Transform)和CCA(Canonical Correlation Analysis)对比方式,10名被试参与此次研究,结果表明HHT解码方式比CCA在准确度方面提高了6.59%;在特征分类方面,采用支持向量机(Support vector machine,SVM)形式。该方向和分级速度控制系统实现了小车在速度和方向上的灵活控制,优化方法提高脑了控小车的准确度与实时性。实验结果显示在选取范式3,刺激时间3s的条件下,10名被试平均识别准确率高达92.50%。文中理论可望为脑控设备走出实验室打下坚实基础。

This paper gives out the design of brain-controlled system of grading speed and direction based on steady-state visual evoked potentials,there are six brain-controlled commands ： le f t, r ig h t, backward,forward （ primary, secondary, te r tia ry）. I t optimized on three aspects ： stimulation paradigm, the stimulation time and spatial layout. The experiments show that expand the stimulus target spacing and strengthen the stimulus time can improve the accuracy of target recognition. In terms of EEG decoding,i t used HHT （ Hilbert-Huang Transform） and CCA （ Canonical Correlation Analysis） contrast mode. Ten subjects participated in the experiment,the results show that the HHT decoding method is 6.59% higher than the CCA in accuracy; in terms of feature classification, support vector machine （SVM） is adopted.The direction and grading speed control system achieved the car flexible control on the speed and direction aspects, optimization methods improved brain-controlled car accuracy and real- time. The experimental results show that the average accuracy of the 10 subjects is 92. 5 0 % under the condition of selecting paragram 3 and stimulus time of 3s. I t is expected for the brain-controlled machine out of the laboratory to lay a solid foundation.