Parametric identification of flexible beam system using evolutionary algorithm

An application of flexible structures in engineering is spread extensively due to lightweight property and technical importance. Before implementing the system, the dynamic behavior of the system needs to be studied by developing a mathematical model. The model-based approach like finite element met...

Full description

Saved in:

Bibliographic Details
Format:	Thesis
Language:	English
Subjects:	Flexible structures Balance beam Structural engineering Beam
Online Access:	http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/2/Full%20text.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/3/Noor%20Fadhilah.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-unimap-78358
record_format	uketd_dc
spelling	my-unimap-783582023-04-17T08:06:47Z Parametric identification of flexible beam system using evolutionary algorithm Mohd Sazli, Saad An application of flexible structures in engineering is spread extensively due to lightweight property and technical importance. Before implementing the system, the dynamic behavior of the system needs to be studied by developing a mathematical model. The model-based approach like finite element method which commonly used in modeling usually required a wide knowledge on the system to be studied and involves complex equations. In system identification technique, the conventional parameter estimation is commonly applied and the limitation is, it may cause the solution trapped in local optima and reduce the efficiency of the model. Therefore, in this study, the model of flexible beam is developed by using system identification method which is based on experimental data collected from the experimental rig and using evolutionary algorithm (EAs) as estimation technique. This research provides a new platform for other researcher to develop a model based on system identification technique using EA’s for other system or application. Other than that, it also delivers a basis for future study on analysis of the performance EAs in terms of different parameter settings in comparison with other algorithms. An attempt of obtaining the linear model is accomplished by developing an experimental rig of flexible beam using square wave signal with mixing resonance frequency to collect input-output data. Auto-regressive with exogenous inputs (ARX) is chosen as a model structure of the system. The coefficient parameters of model structure are estimated via EAs such as firefly algorithm and bat algorithm. A few sets of parameter settings for FA and BA are tested to examine the effect of settings to the performance of model. The best models obtained from each estimation method are compared with least squares algorithm and validated using mean square error (MSE) and one step-ahead prediction (OSA). The main result shows that FA-estimation has MSE of 9.46E-5 which is the lowest among all the estimation method and BA-estimation also outperformed LS-estimation (1.16E-2) by getting lower MSE which is 2.70E-4. Overall of this study proved that evolutionary algorithm able to produce better performance than conventional algorithm. Firefly algorithm and bat algorithm is effective and capable to be used in this area of study like other engineering application Universiti Malaysia Perlis (UniMAP) Thesis en http://dspace.unimap.edu.my:80/xmlui/handle/123456789/78358 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/4/license.txt 8a4605be74aa9ea9d79846c1fba20a33 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/1/Page%201-24.pdf 5256e77959a7ee625dec530579c5ddc1 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/2/Full%20text.pdf 35264b9d02014a03155a04b1f03385a7 http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/3/Noor%20Fadhilah.pdf 616bd668987f0ff38d91e0494d7bb515 Universiti Malaysia Perlis (UniMAP) Flexible structures Balance beam Structural engineering Beam School of Manufacturing Engineering
institution	Universiti Malaysia Perlis
collection	UniMAP Institutional Repository
language	English
advisor	Mohd Sazli, Saad
topic	Flexible structures Balance beam Structural engineering Beam
spellingShingle	Flexible structures Balance beam Structural engineering Beam Parametric identification of flexible beam system using evolutionary algorithm
description	An application of flexible structures in engineering is spread extensively due to lightweight property and technical importance. Before implementing the system, the dynamic behavior of the system needs to be studied by developing a mathematical model. The model-based approach like finite element method which commonly used in modeling usually required a wide knowledge on the system to be studied and involves complex equations. In system identification technique, the conventional parameter estimation is commonly applied and the limitation is, it may cause the solution trapped in local optima and reduce the efficiency of the model. Therefore, in this study, the model of flexible beam is developed by using system identification method which is based on experimental data collected from the experimental rig and using evolutionary algorithm (EAs) as estimation technique. This research provides a new platform for other researcher to develop a model based on system identification technique using EA’s for other system or application. Other than that, it also delivers a basis for future study on analysis of the performance EAs in terms of different parameter settings in comparison with other algorithms. An attempt of obtaining the linear model is accomplished by developing an experimental rig of flexible beam using square wave signal with mixing resonance frequency to collect input-output data. Auto-regressive with exogenous inputs (ARX) is chosen as a model structure of the system. The coefficient parameters of model structure are estimated via EAs such as firefly algorithm and bat algorithm. A few sets of parameter settings for FA and BA are tested to examine the effect of settings to the performance of model. The best models obtained from each estimation method are compared with least squares algorithm and validated using mean square error (MSE) and one step-ahead prediction (OSA). The main result shows that FA-estimation has MSE of 9.46E-5 which is the lowest among all the estimation method and BA-estimation also outperformed LS-estimation (1.16E-2) by getting lower MSE which is 2.70E-4. Overall of this study proved that evolutionary algorithm able to produce better performance than conventional algorithm. Firefly algorithm and bat algorithm is effective and capable to be used in this area of study like other engineering application
format	Thesis
title	Parametric identification of flexible beam system using evolutionary algorithm
title_short	Parametric identification of flexible beam system using evolutionary algorithm
title_full	Parametric identification of flexible beam system using evolutionary algorithm
title_fullStr	Parametric identification of flexible beam system using evolutionary algorithm
title_full_unstemmed	Parametric identification of flexible beam system using evolutionary algorithm
title_sort	parametric identification of flexible beam system using evolutionary algorithm
granting_institution	Universiti Malaysia Perlis (UniMAP)
granting_department	School of Manufacturing Engineering
url	http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/1/Page%201-24.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/2/Full%20text.pdf http://dspace.unimap.edu.my:80/xmlui/bitstream/123456789/78358/3/Noor%20Fadhilah.pdf
_version_	1776104285894868992

Parametric identification of flexible beam system using evolutionary algorithm

Similar Items