Search results

1 – 2 of 2
Article
Publication date: 9 May 2008

Laiq Khan, Tariq Saeed and K.L. Lo

Modern power systems suffer from a well‐known problem of low‐frequency oscillations. Flexible AC transmission systems devices are used to overcome this problem. The aim of…

Abstract

Purpose

Modern power systems suffer from a well‐known problem of low‐frequency oscillations. Flexible AC transmission systems devices are used to overcome this problem. The aim of this paper is to develop a particle swarm optimization (PSO) based supplementary damping control system design for thyristor control series compensator (TCSC).

Design/methodology/approach

The problem is formulated as an optimization problem with an eigenvalue‐based multi‐objective function. PSO is then used to find optimal set of controller parameters by minimizing the objective function.

Findings

The performance and robustness of the proposed approach is validated through small signal and large signal for different loading conditions of a multi‐machine power system.

Originality/value

The paper presents a novel PSO‐based control system design that exhibits robustness and excellent damping performance.

Details

COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, vol. 27 no. 3
Type: Research Article
ISSN: 0332-1649

Keywords

Article
Publication date: 1 October 2006

Laiq Khan, K.L. Lo and S. Jovanovic

The aim of the paper is to develop a novel genetic algorithm (GA)‐based supplementary NeuroFuzzy damping control system for the unified power flow controller (UPFC).

Abstract

Purpose

The aim of the paper is to develop a novel genetic algorithm (GA)‐based supplementary NeuroFuzzy damping control system for the unified power flow controller (UPFC).

Design/methodology/approach

The designed scheme employs a micro‐GA (μ‐GA) to avoid being trapped in a local minimum as opposed to the use of the classical back‐propagation technique. The scheme also uses the “Grand‐Parenting” technique for seeding the initial population to hasten the GA convergence speed. To further speed up the GA for solving the optimization problem, a parallel μ‐GA scheme is also used.

Findings

It has been discovered that a parallel μ‐GA scheme with three computers setup is approximately three times faster than the μ‐GA with a single computer node. Also when μ‐GA is integrated with the “Grand‐Parenting” technique for seeding the initial population, it would hasten the convergence speed. The control scheme exhibits strong robustness and excellent damping performance when tested on a multi‐machine power system.

Originality/value

Presentation of a novel NeuroFuzzy‐based UPFC that exhibits strong robustness and excellent damping performance.

Details

COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, vol. 25 no. 4
Type: Research Article
ISSN: 0332-1649

Keywords

1 – 2 of 2