Font Size: a A A

Visual-Evoked Potential Brain Computer Interface Optimization Algorithm And Its Application

Posted on:2019-08-21Degree:MasterType:Thesis
Country:ChinaCandidate:K J LiuFull Text:PDF
GTID:2370330611993584Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Brain-computer interfaces,billed as transformative,can enable human beings to exchange information with the world by a more efficient and direct way,bypassing the peripheral nerves and muscles.It has helped many severely disabled patients to rebuild their communications with the world.With the technology's continuing development,healthy people also become beneficiaries.Due to the continuous expansion of the application background,it is particularly important to improve the BCI information transmission rate,so that it adapts to frequent,real-time control scenarios.The research in this paper mainly focuses on this goal.First,we propose a two-stage P300 optimization algorithm based on prior information.On the basis of ensuring the correct rate,the stimulus presentation time is effectively shortened,and thus the information transmission rate is greatly improved.It is verified by experiments that compared with the traditional "Oddball" P300 paradigm,our proposed two-stage P300 optimization paradigm based on prior information can increase the information transmission rate by 12.77% to 43.1%,and the average promotion rate is 33%.Next,we continue to optimize the above paradigm.In order to make the two-stage P300 paradigm applicable to the real-life control scenario,we proposed a hybrid braincomputer interface of EOG and P300.We identify three consecutive blinking EOG signals as asynchronous switches.Then P300 was used to provide multiple options for the subject.The experiment determined that the EOG asynchronous switch took an average of 0.93 s to initiate the system,and the false positive rate was 2 times/min.Finally,to meet the intensive control needs in actual control,we propose a shared control framework,and apply the above-mentioned asynchronous fusion paradigm to the simulation of drone cluster control,and initially explore the method of combining human brain intelligence with cluster intelligence.In order to solve the cost imbalance problem in actual control,we add redundancy items that do not trigger any commands in the options,and set a more severe takeoff trigger condition to reduce the false positive rate to zero.
Keywords/Search Tags:Brain-Computer Interfaces, P300, Visual-Evoked Potentials, Optimization Algorithm
PDF Full Text Request
Related items