Construction process identification simulator based on artificial neural networks

Dissertant

Abbud, Zaynab S.

University

University of Technology

Faculty

-

Department

Computer Sciences Department

University Country

Iraq

Degree

Master

Degree Date

2004

English Abstract

Many methods are demanded for identifying processes, with different orders, powerful methods appear to utilize computer software tools to simplify and speed up the identification methods especially for complex systems like nonlinear chemical systems. One of these tools that is covered in this thesis is using artificial neural networks (ANN) software tool, four essential processes were identified using this tool, integrator, dead time, lag, and two-lag processes with three input tested signals impulse step and ramp, then three processes have been built from the essential one, the neural identifiers are NNI (neural network for integrator process), NND (neural network for deadtime process), NNL (neural network for lag process), and NNI, I, (neural network for two-lag process),and the additional processes are NNDI (neural network for deadtime and integrator process), NNDL (neural network for deadtime and lag process), and NNDLL (neural network for deadtime and two-lag process). The present work is achieved by using Levenberg-Marqardet algorithm with variable learning rate, the software used in this work is implemented by using Turbo (C language) with technical MATLAB version 6, 0 package. Time response of the above processes has been achieved by software tools instead of classical hardware circuitry.

These tools are detected as a new approach at the present time.

Main Subjects

Information Technology and Computer Science

Topics

• Artificial intelligence
• Simulation methods
• Data processing
• Logic machines
• Neural networks(Computer science)
• Bionics

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-306025