Design and implementation of fuzzy logic controller for IVAX SCARA robotic manipulator

number: 
3107
English
Degree: 
Author: 
Ayam Mohsen Abbass
Supervisor: 
Dr. Mohammed Z. Al-Faiz
year: 
2013
Abstract:

IVAX Selective Compliance Articulated Robot Arm (SCARA) has a simplified model which is used in industrial areas such as pick and place and automated palletizing.It is used for educational use. A robot allows for 4-degrees of freedom (DOF) in the X-Y plane via two parallel, rotational joints; freedom in the Z-plane via one vertical, linear joint, and the freedom to rotate the end effectors about the Z-axis. It is powered by 4- DC servos with incremental optical encoders providing positional feedback. Two types of controllers are designed and implemented for a robot, they are PID controller and fuzzy like PD controller. The controllers are designed to reach the joints of a robot to desired position with specific orientation. Each joint of the robot is treated as a Single Input Single Output (SISO) system. The response of the first joint (shoulder) of a robot when using Fuzzy Logic Controller (FLC) is smooth and overshoot equal zero but it is slower than when using PID controller. With PID controller, the response contains overshoot and settling time is less than when using FLC. For both controllers the steady state error equal zero. The response when using FLC for the second joint (elbow) of a robot is smooth and overshoot equal zero, it is fast like PID controller. With PID controller, the response contains overshoot and the settling time is less than when using FLC. For both controllers the steady state error equal zero. The response when using FLC for the fourth joint (wrist) of a robot is smooth and overshoot equal zero, it is faster than PID controller. When using PID controller the response contains overshoot and the settling time is less than with FLC. For both controllers the steady state error equal zero.The inverse kinematics (IK) problem of robot manipulator is a complex problem. Because the structure of a robot is simple, one of the simple methods like algebraic approach is used to solve the inverse kinematics problem for a robot. This method is started by finding the angle of the elbow, then finding the angle of the shoulder, the angle of the wrist and the prismatic distance of prismatic joint and it is implemented experimentally for a robot. Mean Square Error (MSE) and objective function of error E (θ) is computed for IK. Two tests of inverse kinematics solutions are implemented for a robot. The values of MSE and E (θ) for the two tests are very small and reach approach zero. The interface card is designed and implemented to connect the robot with the PC computer via multi function card (PCI-1711U), which is supported by Simulink/Real Time Window Target (RTWT) and it will be used for the controller creation and solution. The system programs are built with Simulink/ (RTWT) using Matlab (2011a) which is used to design and implement the simulation of the forward kinematics and inverse kinematics for the robot.