research centers


Search results: Found 5

Listing 1 - 5 of 5
Sort by

Article
Prediction of Fatigue Life of Fiber Glass Reinforced Composite (FGRC) using Artificial Neural Network

Authors: M.S. Abdu_Lateef --- N.S. Abdulrazaq --- A.G. Mohammed
Journal: Engineering and Technology Journal مجلة الهندسة والتكنولوجيا ISSN: 16816900 24120758 Year: 2017 Volume: 35 Issue: 4 Part (A) Engineering Pages: 327-339
Publisher: University of Technology الجامعة التكنولوجية

Loading...
Loading...
Abstract

The present work studies the mechanical properties of composite materials, experimentally and analytically, that are fabricated by stacking 4-layers of fiberglass reinforced with polyester resin. This plies are tested under dynamic load (fatigue test) in fully reversible tension-compression (R=-1) to estimate the fatigue life of the composite where fatigue performance of fiberglass reinforced composed is an increasingly important consideration especially when designing wind turbine blades. In order to predict fatigue life (Number of cycles to failure), conventional analytical techniques are used in the present work. In addition, Artificial Neural Network (ANN) is a reliable and accurate technique that is used for predicting fatigue life. The used networks are; Feed Forward Neural Network (FFNN), Generalized Regression Neural Network (GRNN) and Radial Bases Function Neural Network (RBFNN). Based on the comparison of the results, it is found that the ANN techniques are better than conventional methods for prediction. The results shows that (RBNN2), where stress load and angle of orientation are input to the network and number of cycles to failure as output, is an efficient tool for prediction and optimization the fatigue life of fiberglass reinforced composite.

Keywords

Prediction --- FGRC --- ANN --- FFNN --- GRNN --- RBFNN.


Article
Modified Training Method For Feedforward Neural Networks And Its Application in 4-Link Scara Robot Identification

Authors: Dina A. Abdul Kadeer --- Kais Said Ismail --- Nadia A. Shiltagh
Journal: Journal of Engineering مجلة الهندسة ISSN: 17264073 25203339 Year: 2011 Volume: 17 Issue: 5 Pages: 1335-1344
Publisher: Baghdad University جامعة بغداد

Loading...
Loading...
Abstract

In this research the results of applying Artificial Neural Networks with modified activation function to perform the online and offline identification of four Degrees of Freedom (4-DOF) Selective Compliance Assembly Robot Arm (SCARA) manipulator robot will be described. The proposed model of identification strategy consists of a feed-forward neural network with a modified activation function that operates in parallel with the SCARA robot model. Feed-Forward Neural Networks (FFNN) which have been trained online and offline have been used, without requiring any previous knowledge about the system to be identified. The activation function that is used in the hidden layer in FFNN is a modified version of the wavelet function. This approach has been performed very successfully, with better results obtained with the FFNN with modified wavelet activation function (FFMW) when compared with classic FFNN with Sigmoid activation function (FFS) .One can notice from the simulation that the FFMW can be capable of identifying the 4-Links of SCARA robot more efficiently than the classic FFS

في هذا البحث نتائج تطبيق الشبكات العصبية الصناعية ذات الدالة المحفزة المطورة لتعرف على أداء الروبوت المكون من أربع درجات من الحرية (4 - DoF) لذراع الروبوت (SCARA) سيتم وصفها. النموذج المقترح لإستراتجية التعرف يتكون من شبكة التغذية العصبية ذات الدالة المطورة التي تعمل بالتوازي مع نموذج الروبوت SCARA. تم تدريب الشبكات العصبية ذات التغذية الأمامية (FFNN) على الروبوت ، دون الحاجة إلى أي معرفة سابقة عن النظام المراد التعرف علية.الدالة المحفزة المستخدمة في الطبقة المخفية من الشبكات العصبية الأمامية هي نسخة مطورة من دالة الموجات. وقد نفذ هذا التحوير بنجاح كبير ، مع الحصول على نتائج أفضل عند استخدام FFNN ذات الدالة المحفزة المطورة (FFMW) بالمقارنة مع FFNN الكلاسيكية . من خلال النتائج من الممكن ملاحظة أن FFMW قادرة على تحديد 4 - روابط الى الروبوت نوع SCARA أكثر كفاءة من الشبكات العصبية ذات الدالة المحفزة من نوع Sigmoid


Article
ANN Modified Design Model to Adjust Field Current of D.C. Motor
تصمیم نموذج معدل بالشبكة العصبیة الاصطناعیة لتعدیل تیار الأثارة لمحرك تیار مستمر

Authors: Ahlam Luaibi Shuraiji --- Suad Khairi Mohammed --- Alia Jasim Mohammed
Journal: Engineering and Technology Journal مجلة الهندسة والتكنولوجيا ISSN: 16816900 24120758 Year: 2010 Volume: 28 Issue: 11 Pages: 2132-2142
Publisher: University of Technology الجامعة التكنولوجية

Loading...
Loading...
Abstract

This work is concerned with designing an adjusted field current of D.C.motors to obtain constant speed, based on ANN. The design is employed by usingtraining model with supervised manner with back-propagation algorithm.MATLAB neural network tool box is used for training purpose.The feed-forward neural network (FFNN) and learning capabilities offers apromising way to solve the problem of system non-linearity, parameter variationson unexpected load excisions associated with D.C. motor drive system.The proposed ANN controller model is implemented with a control dc motordrive system in the laboratory. The laboratory test results validate the efficacy ofthe based controller model for a high performance dc motor drive.

يتعلق العمل في هذا البحث بتصميم نموذج يعتمد على تعديل تيار الأثارة لمحركات التيار . ( ANN) المستمر للحصول على سرعة ثابتة، العمل مستند على الشبكة العصبية الاصطناعية استخدم هذا التصميم المقترح مع استعمال خوارزمية التوليد الخلفية لتدريب النموذج بالنسبة للشبكة العصبية أستعمل لأغراض التدريب. (MATLAB) المقترح.صندوق أدوات تعرض الحلول للمشاكل اللاخطية (FFNN) قابليات التعلم للشبكة العصبية ذات التغذية الامامية للمنظومة، اختلافات الباراميترات على الانقطاعات الغير متوقعة للأحمال المرتبطة بمنظومة .( D.C.) محرك تيار مستمر أن نموذج جهاز السيطرة المقترح طبق على منظومة مختبرية لمحرك تيار صدقت النتائج المختبرية كفاءة جهاز السيطرة المستند على الاداء العالي لمحرك .(D.C.)


Article
Intelligent Neural Network with Greedy Alignment for Job-Shop Scheduling

Authors: Fatin I. Telchy1 --- Safanah Rafaat2
Journal: IRAQI JOURNAL OF COMPUTERS,COMMUNICATION AND CONTROL & SYSTEMS ENGINEERING المجلة العراقية لهندسة الحاسبات والاتصالات والسيطرة والنظم ISSN: 18119212 Year: 2015 Volume: 15 Issue: 3 Pages: 11-24
Publisher: University of Technology الجامعة التكنولوجية

Loading...
Loading...
Abstract

Abstract –Job-Shop Scheduling (JSS) processes have highly complex structure interms of many criteria. Because there is no limitation in the number of the process andthere are many alternative scheduling. In JSS, each order that is processed on differentmachines has its own process and process order. It is very important to put theseprocesses into a sequence according to a certain order. In addition, some constraintsmust be considered in order to obtain the appropriate tables.In this paper, a 3-layers Feed Forward Backpropagation Neural Network (FFBNN) hasbeen used for two different purposes, the first one task is to obtain the priority and thesecond one role is to determine the starting order of each operation within a job.Precedence order of operations indicates the dependency of subtasks within a job,Furthermore, the combined greedy procedure along with the back propagation algorithmwill align operations of each job until best solution is obtained. In particular, greedytype algorithm will not always find the optimal solution. However, adding a predefinedalignment dataset along with the greedy procedure result in optimal solutions.


Article
Intelligent Feedback Scheduling of Control Tasks

Author: Fatin I. Telchy
Journal: Iraqi Journal for Electrical And Electronic Engineering المجلة العراقية للهندسة الكهربائية والالكترونية ISSN: 18145892 Year: 2014 Volume: 10 Issue: 2 Pages: 64-79
Publisher: Basrah University جامعة البصرة

Loading...
Loading...
Abstract

an efficient feedback scheduling scheme based on the proposed Feed Forward Neural Network (FFNN)scheme is employed to improve the overall control performance while minimizing the overhead of feedbackscheduling which exposed using the optimal solutions obtained offline by mathematical optimization methods. Thepreviously described FFNN is employed to adapt online the sampling periods of concurrent control tasks withrespect to changes in computing resource availability. The proposed intelligent scheduler will be examined withdifferent optimization algorithms. An inverted pendulum cost function is used in these experiments. Then,simulation of three inverted pendulums as intelligent Real Time System (RTS) is described in details.Numerical simulation results demonstrates that the proposed scheme can reduce the computationaloverhead significantly while delivering almost the same overall control performance as compared to optimalfeedback scheduling

Listing 1 - 5 of 5
Sort by
Narrow your search

Resource type

article (5)


Language

English (5)


Year
From To Submit

2017 (1)

2015 (1)

2014 (1)

2011 (1)

2010 (1)