CONTROL OF HVDC POWER SYSTEM USING ARTIFICIAL INTELLIGENCE

MOHAMMED MAHMOUD MOHAMMED AFIFI;

Abstract


High voltage direct current (HVDC) transmi ssion controllers may not be optimi zed for a wide range of operating conditions because they ex hi bit a fi xed gain structure.
A feasible altern ati ve proposed to overcome this li mitation is the use of an Artifi cial Neura l Networks (ANN) controller. This thesis describes an applicati on of ANN control technique in II V DC transm ission lin k for d ynamic performance improvement.

The proposed ANN utili zes the training data of the controllers optimal gains obtained from the time doma in simulation using the Pl fixed gai ns conventiona l controll er structure combined with gain scheduling technique for stead y state and fau l t conditions.

Two cases of disturbances are considered to evaluate the performance of the ANN, compared w ith the con ventional P I controller, in HVDC power control.
First is the small signal disturbance su ch as t he step change in t he cu rrent regula tor reference valu e and sudden load cha nge (increase or decrease). Second is the large signal disturbance such as three-phase short circuit at the genera tor bus bar.


The MATLAB and the NEURAL NETWORKS TOOLBOX software are used for conventional and ANN controller simul ation. The transient simulation results a re obtained and conclusi ons are drawn.


Other data

Title CONTROL OF HVDC POWER SYSTEM USING ARTIFICIAL INTELLIGENCE
Other Titles التحكم في نقل القدرة الكهربيائية عبر خطوط الجهد العالي المستمر باستخدام الذكاء الصناعي
Authors MOHAMMED MAHMOUD MOHAMMED AFIFI
Keywords Artifi cial Neural Networks (ANN), Traini ng Data, HVDC Transmi ssion, Gain Sched uling, and Con t rol Sensiti v it y Index.
Issue Date 1999

Attached Files

File SizeFormat
MOHAMMED MAHMOUD MOHAMMED AFIFI.pdf2.09 MBAdobe PDFView/Open
Recommend this item

Similar Items from Core Recommender Database

Google ScholarTM

Check



Items in Ain Shams Scholar are protected by copyright, with all rights reserved, unless otherwise indicated.