Hasbulah, Muhammad Haziq (2021) Implementation of EEG controlled technology to modular self-reconfigurable robot with multiple configurations. Masters thesis, Universiti Teknikal Malaysia Melaka.
![[img]](http://eprints.utem.edu.my/style/images/fileicons/text.png) |
Text (24 Pages)
Implementation of EEG controlled technology to modular self-reconfigurable robot with multiple configurations.pdf - Submitted Version Download (445kB) |
|
Text (Full Text)
Implementation of EEG controlled technology to modular self-reconfigurable robot with multiple configurations.pdf - Submitted Version Restricted to Registered users only Download (9MB) |
Abstract
The electroencephalogram (EEG) implementation has reached a new level in terms of application that is for the Brain Computer Interfaces (BCI) system and not restricted for medical instrumentation only. The concept of Modular self- Reconfigurable (MSR) robot control can be identified in most of science fictional movies. The implementations of both technologies to each other will act as a frontier for new alternatives that improve self-reconfigurable modular robots in terms of the control strategy. The main problem is that the EEG-based BCI system is always implemented for mobile robots, robot manipulators, and sometimes on humanoid robots. However, it is not implemented to MSR robots, which perform their tasks cooperatively by more than one robot module. Hence, the EEG-based BCI system implementation to MSR robot is needed to ensure the high accuracy of the MSR robot controlled with the BCI system to assess multiple configuration propagations by the MSR robots regardless of external stimulation. Therefore, it is important to analyse society perspective on BCI controlled robot technologies, to establish control, and to assess multiple configurations propagate by the Dtto MSR robot based on the EEG-based BCI system. Finally, the system established needs to be analyzed in terms of versatility for the availability of training, gender, and robot state. The method proposed in our study is utilizing Lab Streaming Layer (LSL) and Python script as mediators. The system developed in our study was done by using OpenViBE software where a Motor Imagery BCI was created to receive and process the EEG data in real time. The main idea for the developed system is to associate a direction (Left, Right, Up, and Down) based on Hand and Feet Motor Imagery as a command for the Dtto MSR robot control. Based on the findings, the SVM classifier produces a better result for Motor Imagery system control accuracy. The study also shows that the EEG acquisition headset with multiple electrodes is necessary for achieving a better control accuracy for the Motor Imagery system. A deeper analysis of the versatility of the MSR robot controlled by the BCI system is based on the three factors that were decided. Highest success rate for Simulation based on Left imagery which is 27.5% and the highest success rate for Young Trained subjects which is 30%. Highest success rate for Real Robot based on Left and Right Imagery which is 37.5% and the highest success rate for Young Trained subjects which is 38.33%. The analysis result shows that “Aged” and “Robot State” are significant for the control success rate of MSR robots by the BCI system. As for the “Training Availability” factor in our study, it is not considered a significant factor on its own but it has an interaction with the other factors and influences the control success rate. Overall, it is something achievable as the BCI system was integrated to the MSR robot to control multiple robot modules in real time and produced positive result as intended even though it was not as high as expected. P300 or SSVEP brain signal can be implemented in the future for more degree of freedom control and more efficient way can be implemented for communication for BCI system to MSR robot.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Brain-computer interfaces, Robots, Control systems, Human-machine systems |
| Subjects: | T Technology > T Technology (General) T Technology > TJ Mechanical engineering and machinery |
| Divisions: | Library > Tesis > FKP |
| Depositing User: | F Haslinda Harun |
| Date Deposited: | 26 Jan 2023 17:00 |
| Last Modified: | 24 Feb 2023 08:07 |
| URI: | http://eprints.utem.edu.my/id/eprint/26066 |
| Statistic Details: | View Download Statistic |
Actions (login required)
 |
View Item |