Vibration control of single link flexible manipulator by using neural network

This project presents simulation on minimizing vibration error in single link flexible manipulator system by using neural network system. Flexible manipulator system has a flexible link, an actuator-gear mechanism to rotate the link, an optical encoder to measure joint rotation, accelerometers and s...

Full description

Saved in:
Bibliographic Details
Main Author: Muhammad Zulhilmi, Zakaria
Format: Undergraduates Project Papers
Language:English
Published: 2013
Subjects:
Online Access:http://umpir.ump.edu.my/id/eprint/8582/1/CD8055_%40_71.pdf
http://umpir.ump.edu.my/id/eprint/8582/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This project presents simulation on minimizing vibration error in single link flexible manipulator system by using neural network system. Flexible manipulator system has a flexible link, an actuator-gear mechanism to rotate the link, an optical encoder to measure joint rotation, accelerometers and strain gauges to sense flexible motion, an optical arrangement to measure the endpoint position and an occasional force sensor attached to the end-point. The overall aim of this project is to develop a dynamic modeling and controller for single link flexible manipulator. In spite of it, we need to minimize the vibration using neural network controller in single link flexible manipulator. The vibration error that occurs in the flexible manipulator is needed to be study and try to reduce it by using the controller (neural network). Towards this thesis, the single link flexible manipulator system being minimize the error by using intelligent neural network controller, and be compared with system existing controller (PID) and the system without controller so that we can see the clearly error percentage reduced. In order to achieve the objective for this project, mathematical model will develop based on system identification using different method such as Lagrage method, Euler-Beurnoulli and System Identification Toolbox in MATLAB and implement it in Mathlab simulink. The results that we achieved is the neural network give the best in order to minimize the vibration error compared to the system without controller about 60% reduction of error and 10% of reduction of error when compared with system with PID controller. Conclusively, the intelligent neural network give us the better results and followed the characteristic of single link flexible manipulator as we desired.