Constrained spectral clustering‐based methodology for intentional controlled islanding of large‐scale power systems

Abstract

Abstract:Intentional controlled islanding is an effective corrective approach to minimise the impact of cascading outages leading to large-area blackouts. This paper proposes a novel methodology, based on constrained spectral clustering, that is computationally very efficient and determines an islanding solution with minimal power flow disruption, while ensuring that each island contains only coherent generators. The proposed methodology also enables operators to constrain any branch, which must not be disconnected, to be excluded from the islanding solution. The methodology is tested using the dynamic models of the IEEE 39-and IEEE 118-bus test systems. Time-domain simulation results for different contingencies are used to demonstrate the effectiveness of the proposed methodology to minimise the impact of cascading outages leading to large-area blackouts. In addition, a realistically sized system (a reduced model of the Great Britain network with 815 buses) is used to evaluate the efficiency and accuracy of the methodology in large-scale networks. These simulations demonstrate that our methodology is more efficient, in a factor of approximately 10, and more accurate than another existing approach for minimal power flow disruption.

Intentional Controlled Islanding and Risk Assessment: A Unified Framework

“…The Concept of ICI Fig. 1 illustrates the general concept of ICI [9], which is associated with the blackout progress [21]. Blackouts might be caused by the uncontrolled separation of the system [1], [2].…”

“…The slow coherency theory is effectively used in [4]- [6] to split the system across weak connections; however, this approach might not account for changes in the topology of the network, which can have an effect in the coherency results. Spectral clustering-based ICI methods are presented in [7]- [9]. However, the unsupervised nature of these approaches makes difficult the introduction of ICI constraints (e.g., generator coherency, load-generation balance) in the problem.…”

“…Ordered binary decision diagrams (OBDD) are also used [10]- [12]. Although OBDD-based ICI methods can search for a solution across the whole search space, the size of the network to be partitioned needs to be reduced to less than 40 nodes to obtain solutions within realistic time frames (i.e., seconds) [9].…”

“…1. General framework of the intentional controlled islanding [9] tiveness. Multiple scenarios in the real system are developed to demonstrate its adaptability to different realistic conditions.…”

Two-stage controlled islanding strategy based on Stoer-Wagner and improved Dinic algorithms

“…The clustering algorithm employs the relaxation methods so as to broaden constraints, which needs to be verified in large grids [17]. Reference [18] proposed an improved constrained spectral clustering algorithm to obtain an optimal islanding result. However, the computational efficiency needs be improved for the large girds, and the generator coherency is not taken into account.…”