“Optimal and Flexible Strategy for Adaptive Control of Active Filters Under Non-Sinusoidal Line Voltages”
 
 

Advisors :
1. Dr. Reza Ghazi, Dept. of Electrical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
2. Dr. Hamid A. Toliyat,  Dept. of Electrical Engineering, Texas A&M University , Texas, USA.
 

Abstract :

    This Dissertation gives a new insight into the concept of load compensation under distorted voltages.  Achieving both unity power factor (UPF) and perfect compensation of current harmonics are not possible where a competition will arise between these two important factors.  Through evaluating the present  control strategies, a  generalized, optimal, and flexible control strategy (OFC) for harmonic compensation of utility lines is proposed. The proposed control strategy, which provides a unified and highly flexible compensation framework has the ability of programming for perfect current harmonics compensation, or (UPF) accomplishment, or other newly defined objectives such as maximizing the power-factor subject to some adjustable constraints on the level of current harmonics and unbalancing via an on-line optimization algorithm.  The strategy can fulfill the IEEE-519 standards requirements, while guaranteeing the best achievable power factor and optimum  required rating for the compensator.  A fast and simple version of OFC strategy has been  successfully proposed  and realized using neural networks . The optimization algorithm  is replaced by a neural network based system  that is learned off-line  and has the real time realization capability .Theoretical concepts and practical features of the proposed control strategies have been shown through extensive simulation studies using MATLAB/SIMULINK  programs.

Key Words: Distorted voltage, harmonics, power factor, active filter, flexible compensation strategy,  optimization, real –time control, neural networks .