基于改进单关节信息传输模型的闭环脑机接口系统设计

Closed-loop brain-machine interface system design based on improved single-joint information transmission model

近年来,脑机接口(BMI)技术在残疾人肢体功能康复、老年人生活辅助等方面的应用日益广泛.本文以单关节信息传输(SJIT)模型为对象,通过模型改进、设计解码器和辅助控制器构造了闭环脑机接口系统以恢复单关节的运动功能.本文主要工作包括:(1)引入相对速度向量对单关节信息传输模型改进以降低模型输出的超调量,并测试了改进模型的性能;(2)基于改进模型,通过设计基于维纳滤波的解码器、基于预测控制策略的辅助控制器构造了闭环脑机接口系统以恢复缺失的信息通路.离线和在线仿真说明,改进模型的输出性能有较大提升、超调量明显下降;构建的闭环系统很好地实现了对缺失信息通路的恢复和期望轨迹的跟踪,且具有较强的抗干扰性.

In recent years,brain-machine interface(BMI)technology has been used more and more widely in physical rehabilitation and life support for the disabled and the elderly.For the the single-joint information transmission(SJIT)model,through the model improvement,the decoder design and the auxiliary controller design,the closed-loop BMI system is formulated in this paper to restore the movement of the single joint.The innovation of this paper mainly includes:(1)the relative velocity vector is introduced to improve the SJIT model to reduce the overshoot,and then the performance of improved model is tested;(2)based on this improved model,a decoder based on Wiener filter and an auxiliary controller based on model predictive control strategy are designed and introduced to restore the missing information loop.The offline and online simulation results show that the improved model can greatly improve the output performance,reduce the overshoot clearly;and the formulated closed-loop system can well restore the missing information loop and track the target trajectory.In addition,the closed-loop system has strong anti-interference capability.