Design & Implementation of a Control Unit for a Tongue Control System
Studenteropgave: Kandidatspeciale og HD afgangsprojekt
- Jeremy Leresteux
- Umberto Cerasani
10. semester, Elektronik og IT, Kandidatuddannelsen (Spec. Applied Signal Processing and Implementation) (Kandidatuddannelse)
This 10th semester project for the «Applied Signal Processing and Implementation» specialization at Aalborg University is a study on the improvement and implementation of the algorithms of a Control Unit for a Tongue-based Control System. The project focuses on a medical application permitting motion-disabled people to control their environment, i.e. external devices like computers, wheelchair, mobile phones,...
The Tongue Control System is developed by the HST Department at Aalborg University where prototypes have been designed. The Mouthpiece Unit (MU), which receives the commands from the user by means of his tongue, is fully developed. This unit sends the different instructions to a Central Unit (CU) for the control of the external devices.
This project seeks, firstly, to improve the algorithms for processing the signals provided by the MU. Indeed, because of its characteristics, noise and drifts in temperature are added to the useful signal. After an analysis of both signal perturbations, the successful custom Kalman filter removes these pertubations without any delay or data loss. The second part of the project deals with the analysis of different platforms to choose the appropriate one for the implementation of the communication and interfaces algorithms. The project group chooses the FPGA Altera DE2 because its flexibility ensures easy additions of future interfaces for many external devices. Moreover, the Kalman filter is implemented on this new platform. The third part explains how to use the tools provided with the Altera DE2 and the results obtained on the implementation. The algorithm performing the logic to control a computer mouse cursor is successfully implemented but the core of the project is the work of this mouse cursor control with the Kalman filter. The process of the signal from 8 sensors, has been successfully implemented and gives good results. Combined with the mouse cursor control, it is possible to control the mouse cursor from the MU and produce smooth and accurate moves. Finally, the conclusion and discussion about the performances are developed in the last part, as well as the future perspectives discussions.
Sprog | Engelsk |
---|---|
Udgivelsesdato | 2009 |
Antal sider | 103 |
Udgivende institution | AAU |
Emneord | Implementation, Signal processing, Medical application, Kalman, FPGA |
---|