Denoising EMG signal and its implementation for a robot arm using 8051 microcontroller

Dissertant

Ahmad, Anas Fuad

Thesis advisor

Abbas, Iyad Ibrahim
Rajab, Khalidah S.

University

University of Technology

Faculty

-

Department

Department of Electrical Engineering

University Country

Iraq

Degree

Master

Degree Date

2013

English Abstract

Electromyographic (EMG) signal is the electrical activation signal for the muscle, which in turn enables the human arm to do certain movements.

The SEMG signal acquires noise during its recording and detection.

This research consists of two main phases: First phase: In this phase an algorithm based on wavelet transforms is proposed for denoising two SEMG signals captured from two muscles of the human arm (biceps brachii and abductor pollicis brevis).

This algorithm includes nominating five wavelet filters, which are very close to the shape of the original SEMG signals and it aims to find the optimal wavelet filter among them and the optimal decomposition level for denoising these signals.

Also includes, processing these signals to make them suitable for controlling the robotic arm.

The experimental results showed the fourth order daubechies (db4) is the optimal wavelet filter for denoising the SEMG signal captured from biceps brachii muscle and the optimal decomposition level is four.

On the other side the eighth order Symlet (sym8) is the optimal wavelet filter for denoising the SEMG signal captured from abductor pollicis brevis muscle and the optimal decomposition level is four.

This phase was implemented using MATLAB (R2008a) program.

Second phase: This phase includes designing and implementing a robotic arm with two degrees of freedom (2-DOF).

The robotic arm was designed from aluminium tubes with rectangular cross section and the plastic material.

The stepper motors, helical gears, worm gear and belt are used in order to enable the robotic arm for moving at certain angles.

Also, this phase includes designing and implementing control system to control the movements of the robotic arm.

The control system consists of a microcontroller (8051) and interfacing circuit between microcontroller and robotic arm.

The proposed system design can operate in two modes: manual mode to enable the user to interact directly with the robotic arm, and the real SEMG signal mode where in this mode the microcontroller receives the contraction and relaxation levels of the SEMG signals that processed in the first phase via USB-RS232 adapter to control the robotic arm, this mode basically depends on the linear relationship between the level of muscle contraction and the mechanical movement resulting from it.

The proposed design is characterized by the possibility of development and modification.

Main Subjects

Electronic engineering

Topics

• Computing
• Robots

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-417592