This study explored methods for improving the performance of Steady-State Visual Evoked Potential(SSVEP)-based Brain-Computer Interfaces(BCI), and introduced a new analytical method to quantitatively analyze and refle...This study explored methods for improving the performance of Steady-State Visual Evoked Potential(SSVEP)-based Brain-Computer Interfaces(BCI), and introduced a new analytical method to quantitatively analyze and reflect the characteristics of SSVEP. We focused on the effect of the pre-stimulation paradigm on the SSVEP dynamic models and the dynamic response process of SSVEP, and performed a comparative analysis of three pre-stimulus paradigms(black, gray, and white). Four dynamic models with different orders(second-and third-order)and with and without a zero point were used to fit the SSVEP envelope. The zero-pole analytical method was adopted to conduct quantitative analysis on the dynamic models, and the response characteristics of SSVEP were represented by zero-pole distribution characteristics. The results of this study indicated that the pre-stimulation paradigm affects the characteristics of SSVEP, and the dynamic models had good fitting abilities with SSVEPs under various types of pre-stimulation. Furthermore, the zero-pole characteristics of the models effectively characterize the damping coefficient, oscillation period, and other SSVEP characteristics. The comparison of zeros and poles indicated that the gray pre-stimulation condition corresponds to a lower damping coefficient, thus showing its potential to improve the performance of SSVEP-BCIs.展开更多
快反镜作为高精度光学系统的核心执行机构,其控制策略的优劣直接影响系统性能。传统的快反镜控制策略依赖独立的控制盒和线性功放模块,导致快反镜控制系统体积大、热量高、应用受限。而基于金属-氧化物半导体场效应晶体管(metal-oxide-s...快反镜作为高精度光学系统的核心执行机构,其控制策略的优劣直接影响系统性能。传统的快反镜控制策略依赖独立的控制盒和线性功放模块,导致快反镜控制系统体积大、热量高、应用受限。而基于金属-氧化物半导体场效应晶体管(metal-oxide-semiconductor field-effect transistor,MOSFET)开关量驱动的快反镜在控制上难以达到精度要求。通过对快反镜的建模扫频分析,提出了开环零极点配置与比例积分控制(proportional integral derivative,PID)相结合的控制策略,并在快反镜的硬件平台上进行了仿真验证。实验结果表明,该控制策略在不依赖额外控制器和驱动模块的前提下,能够保证快反镜控制的高精度要求。展开更多
基金supported by the Key Research and Development Program of Guangdong Province (No. 2018B030339001)the National Key Research and Development Program of China (No. 2017YFB1002505)the National Natural Science Foundation of China (No. 61431007)
文摘This study explored methods for improving the performance of Steady-State Visual Evoked Potential(SSVEP)-based Brain-Computer Interfaces(BCI), and introduced a new analytical method to quantitatively analyze and reflect the characteristics of SSVEP. We focused on the effect of the pre-stimulation paradigm on the SSVEP dynamic models and the dynamic response process of SSVEP, and performed a comparative analysis of three pre-stimulus paradigms(black, gray, and white). Four dynamic models with different orders(second-and third-order)and with and without a zero point were used to fit the SSVEP envelope. The zero-pole analytical method was adopted to conduct quantitative analysis on the dynamic models, and the response characteristics of SSVEP were represented by zero-pole distribution characteristics. The results of this study indicated that the pre-stimulation paradigm affects the characteristics of SSVEP, and the dynamic models had good fitting abilities with SSVEPs under various types of pre-stimulation. Furthermore, the zero-pole characteristics of the models effectively characterize the damping coefficient, oscillation period, and other SSVEP characteristics. The comparison of zeros and poles indicated that the gray pre-stimulation condition corresponds to a lower damping coefficient, thus showing its potential to improve the performance of SSVEP-BCIs.
文摘快反镜作为高精度光学系统的核心执行机构,其控制策略的优劣直接影响系统性能。传统的快反镜控制策略依赖独立的控制盒和线性功放模块,导致快反镜控制系统体积大、热量高、应用受限。而基于金属-氧化物半导体场效应晶体管(metal-oxide-semiconductor field-effect transistor,MOSFET)开关量驱动的快反镜在控制上难以达到精度要求。通过对快反镜的建模扫频分析,提出了开环零极点配置与比例积分控制(proportional integral derivative,PID)相结合的控制策略,并在快反镜的硬件平台上进行了仿真验证。实验结果表明,该控制策略在不依赖额外控制器和驱动模块的前提下,能够保证快反镜控制的高精度要求。
文摘传统基于离线模型参数和典型运行方式设计的电力系统阻尼控制器存在适应性问题,提出一种基于辨识的自适应控制器设计方法,可解决一般自适应控制中快速性和准确性要求之间的矛盾。所用的多元自回归滑动平均模型(auto regressive moving averaging vector,ARMAV)辨识在电网正常运行过程中针对由负荷等随机扰动引起的类噪声信号进行;在综合考虑辨识误差、阻尼要求和稳定裕度基础上,提出阻尼控制零极点配置基本原则,并设计相应的遗传算法优化方法。为了充分检验上述辨识与控制系统的效果,基于广域测量平台对其进行软硬件实现,并在东北电网简化系统中进行实时数字仿真(real time digital simulation,RTDS)测试,实验结果说明了所提方法的可行性和有效性。
