文献类型：会议论文

英文题名：Improvement of Brain - Machine Interface Model on Dynamic Reaching Task

作者：Sun, Jinggao[1];Xue, Rui[1];Pan, Hongguang[2];Yang, Jiaxiong[1];Wang, Shuo[1]

机构：[1]East China Univ Sci & Technol, Minist Educ, Key Lab Adv Control & Optimizat Technol, Shanghai 200237, Peoples R China;[2]Xian Univ Sci & Technol, Sch Elect & Control Engn, Xian 710049, Shaanxi, Peoples R China

会议论文集：37th Chinese Control Conference (CCC)

会议日期：JUL 25-27, 2018

会议地点：Wuhan, PEOPLES R CHINA

语种：英文

外文关键词：Brain-Machine Interface; Neural Activity Model; Weiner Decoder

摘要：Based on the relative and required velocity integration to endpoint (RRVITE) model, we extend the neuroprosthetic model for voluntary single joint reaching task by adding the vector models related to relative speed information, which improves the model performance in dynamic reaching task. Through the synthetic simulation, we show that the extended model performs much better than the original one in motion prediction and steady state accuracy. Using synthetic data obtained through the simulation of the original model, a linear decoder based on Weiner filter has been trained and introduced into the improved model to test its performance. Both the evaluation results and the online test results show that the improved model can also enhance the performance of the BMI system with decoder.