Analysis ofinduction heating system based on artificial neural networks

Dissertant

Ali, Muhammad Munis Izz al-Din

University

University of Technology

Faculty

-

Department

Electronic Engineering Department

University Country

Iraq

Degree

Ph.D.

Degree Date

2008

English Abstract

Induction heating system has a number of inherent benefits compared to traditional heating systems.

Many analytical and numerical approaches have been applied to solve the problem of induction heating, efforts are made to introduce new development in computational approaches and utilization of newly developed specialized software packages to improve the methods of analysis and design of induction heating systems. Artificial Neural Networks (ANNs) possess many advantages and they have the ability to tackle problems that cannot be accomplished by more analytical and numerical methods.

This work attempts to mix the advantages of both Finite Element Method (FEM), and artificial neural networks.

The first stage of this work includes developing a general and comprehensive program, which works under ANSYS 11.0SP1 package environment, and can be applied to analyze a wide range of induction heating applications.

The next stage involves modeling many artificial neural networks, and training them based on the results of the first stage to enable them to evaluate the heat distribution inside the workpiece of any induction heating system, also neural networks are used to specify the time and the power supply required for any desired heat distribution inside the workpiece with high accuracy and nil run time.

The neural networks are simulated by using Neural Network Toolbox in MATLAB (R2007a), and the networks are trained, according to supervised scaled conjugate gradient algorithm, until the performance function (mean square error) reach the goal (=10-4). The combination of FEM and ANNs, in solving the problem of induction healing shows a good success thro obtaining results with high accuracy and very short run time.

Main Subjects

Electronic engineering

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-305031