Chu, Timothy and Chua, Alvin and Secco, Emanuele Lindo (2022) A Study on Neuro Fuzzy Algorithm Implementation on BCI-UAV Control Systems. ASEAN Engineering Journal. ISSN 2586-9159 (Accepted for Publication)
Preview |
Text
[Revised]A STUDY ON NEURO FUZZY ALGORITHM IMPLEMENTATION ON BCI-UAV CONTROL SYSTEMS.pdf - Accepted Version Download (594kB) | Preview |
Abstract
Brain-Computer Interface (BCI) machines are capable of obtaining brain activities by conducting Electroencephalogram tests. Developments on both BCI and Machine Learning allowed various researchers to develop and study various BCI control systems, mainly varying with the algorithm implementation.
This research presents a performance analysis of the Adaptive Neuro-Fuzzy Inference System (ANFIS) for BCI control systems for drone maneuverability. Eye gestures were used to generate the EEG data that were captured using the Emotiv INSIGHT Neuroheadset. The obtained data were transferred to the computing hardware using IEEE 802.15 wireless communication protocol (i.e. Bluetooth connectivity); the data are processed using the 5th order Butterworth Band-Pass filtering and heuristic filtering. The filtered dataset is then fed to the ANFIS and a Support Vector Machine (SVM) algorithm, the latter serving as the basis, for training and quadcopter control implementation.
Three flight tests were done, hover test, flight command test, and the flight control test to obtain the performance of the control system in terms of accuracy. Results from the initial two tests showed that the ANFIS performed comparably with the SVM, and even about 2% better with an accuracy of 79%. The final test showed that the BCI control system had a maximum variance of 4% compared to the handheld remote controller, where the latter served as the basis. It was found that between Machine Learning algorithms, ANFIS is as capable as the SVM for BCI control systems. Further developments may focus on employing time-series EEG preprocessing techniques.
| Item Type: | Article |
|---|---|
| Additional Information and Comments: | Open Access Publication. This is the author's version of an article that has been accepted for publication in Asean Engineering Journal. The final, published version will be available from https://journals.utm.my/index.php/aej |
| Faculty / Department: | Faculty of Human and Digital Sciences > Mathematics and Computer Science |
| Depositing User: | Emanuele Secco |
| Date Deposited: | 21 Jan 2022 09:26 |
| Last Modified: | 21 Jan 2022 09:28 |
| URI: | https://hira.hope.ac.uk/id/eprint/3477 |
Actions (login required)
 |
View Item |
Altmetric
Altmetric
