Large-scale Power Networks Research Articles

Incorporation of a UPFC model in an optimal power flow using Newton's method

The paper addresses the issue of unified power flow controller (UPFC) modelling within the context of optimal power flow (OPF) solutions. The nonlinear optimisation problem is solved by Newton's method leading to highly robust iterative solutions even for cases of large-scale power networks, where hundreds of variables are to be optimised. It is shown in the paper that networks modified to include several UPFCs are solved with equal reliability. The UPFC model itself is very flexible, it allows the control of active and reactive powers and voltage magnitude simultaneously. It can also be set to control one or more of these parameters in any combination or to control none of them. Considerable progress has been achieved in UPFC modelling intended for conventional load flow studies but this is the first time that the more complex issue of UPFC modelling intended for OPF solutions has been addressed.

Fast calculation of a voltage stability index of power systems

The minimum singular value of the Jacobian matrix of the load flow equation may have been preferred as an indicator of voltage collapse when the static voltage stability of power systems is studied. In this paper, the authors propose a highly efficient algorithm to calculate the smallest singular value of a Jacobian matrix of the load flow equation by employing the noniterative characteristic of an incremental condition estimation (ICE) method and the sparsity characteristic of large scale power networks. Both theoretical bases and computation costs of the algorithm are also detailed in the context. Finally, a practical application example is also given for demonstration.

95/00471 Application of partitioning techniques for decomposing large-scale electric power networks

95/00471 Application of partitioning techniques for decomposing large-scale electric power networks

Application of partitioning techniques for decomposing large-scale electric power networks

Two efficient heuristic algorithms for solving cluster problems associated with partitioning of power networks are presented in this paper. Both algorithms are divided into two stages. In the first stage, an initial partition is created based on the electrical distance among system buses. The second stage involves interchanging pairs of buses among the various clusters of the initial partition. The first algorithm solves an optimal k-decomposition problem. In the optimal k-decomposition, the partitioning of a network into k clusters of buses is performed by maximizing the number of uncut links among clusters. This type of decomposition is known to be equivalent to a 0–1 quadratic programming problem which is approximated by a linear transportation problem. By defining a link weight as the electrical distance between linking buses, the algorithm clusters strongly-connected buses together while weakly-connected buses are placed in different clusters. The second algorithm solves the placement problem of n-connected buses in a k-dimensional Euclidean space; such a problem is reduced to finding k eigenvectors of a connectivity matrix, defined as the bus admittance matrix. The node interchange is an iterative heuristic method that can be used to improve an initial partition, such as that obtained by either the k-decomposition technique or the eigenvector approach. The method moves one bus at a time, from one cluster of the initial partition to another, in an attempt to maximize the total electric distance among clusters of the final partition. Applications of the proposed algorithms to both small and medium-size power systems, and comparing the results with those obtained in another study, are illustrated in this paper.

Linear reactive power optimization in a large power network using the decomposition approach

A mathematical framework is presented for the solution of the economic dispatch problem. The application of the Dantzig-Wolfe decomposition method for the solution of this problem is emphasized. The system's optimization problem is decomposed into several subproblems corresponding to specific areas in the power system. The upper bound technique along with the decomposition method are applied to a 16-bus system and a modified IEEE 30-bus system, and numerical results are presented for larger systems. The results indicate that the presented formulation of the reactive power optimization and the application of the decomposition procedure will facilitate the solution of the problem. The algorithm can be applied to a large-scale power network, where its solution represents a significant reduction in the number of iterations and the required computation time. >

An Optimal Set of Measurements for the Estimation of System States in Large-Scale Power Networks

Abstract This paper addresses the problem of measurement selection and meter locations as related to the power system state estimation. The measurement redundancy is selected as a criterion, and based on the redundancy identification technique applied to linearly constrained mathematical programming problems, a method is proposed to identify redundant measurements among a given set of measurements. Such an identification will determine an adequate set of measurements for state estimators that adopt mathematical programming techniques. The method is rather simple, yet it significantly contributes towards the solution of the measurement selection problem and can be implemented off-line and on real-size systems.

Co-Ordinated Self-Tuning Control for Turbine and Exciter of a Turboalternator

ABSTRACT Multivariable minimum variance self-tuning control of the exciter voltage and turbine speed of an isolated a–c turbogenerator system connected to a large-scale power network is considered. Simulation studies, employing a high-order non-linear representative system model are reported, with emphasis on rigorous assessment of performance within an operational context. Three-phase faults, with and without permanent line-switching, dynamic and transient stability, are amongst problems addressed, and comparison with conventional control philosophies is made. Convergence of the self-tuning algorithm under a wide range of operating conditions is demonstrated, and results indicate that a self-tuning controller, based on minimal advance plant knowledge, provides a viable, robust alternative to conventional schemes within the power system environment.

Reactive power control for real power-loss minimization

An accurate and fast technique for real-time control of system voltage and reactive power of large-scale power systems is presented. The one-bus-away method has been modified to obtain accurate results with moderate computer requirements for online application to large-scale power networks. The approach was used to obtain a constrained minimum of the objective function for each subsystem in a deterministic number of steps, in most cases using the special geometric features of the problem. For the few cases where the optimal solution was not obtained from direct steps, good initial starting conditions for the iterative nonlinear optimal technique were obtained. The technique used was the flexible tolerance method, which works directly on objective functions and on inequality constraints without evaluating derivatives or gradients. It was selected because of its robustness, ease of implementation, and fast convergence properties for a wide range of nonlinear small-size problems. The required CPU time was small enough for significant online applications.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>