A topology assessment system for power systems using active power measurements as input data is presented. The method is designed to be incorporated into a state estimator working with a bus-branch oriented network model. The system architecture contains two stages: (i) the preprocessing stage; and (ii) the classification stage. The preprocessing stage transforms each current measurement set to produce a vector in the [0,1]n space. This stage produces clusters of very similar preprocessing output vectors for each grid topology. The classification stage consists in a layer of Gaussian potential units with Mahalanobis distance, and classifies the preprocessing output vectors to identify the actual topology. The main features of this method are: (i) local topology identification; (ii) linear growth of the complexity with the power system size; (iii) correction of multiple errors; and (iv) insensitivity to bad data. Tests have been carried out using the IEEE 14, 30, 57, 118 and 300 standard networks and different topological and measurement configurations. These tests have demonstrated the successful application of the technique.
[1]
S. C. Srivastava,et al.
Topology processing and static state estimation using artificial neural networks
,
1996
.
[2]
K. Clements,et al.
Detection and identification of topology errors in electric power systems
,
1988
.
[3]
H. Glavitsch,et al.
Detection and identification of topological errors in online power system analysis
,
1991
.
[4]
A. M. Leite da Silva,et al.
Online topology determination and bad data suppression in power system operation using artificial neural networks
,
1997,
Proceedings of the 20th International Conference on Power Industry Computer Applications.
[5]
I. S. Costa,et al.
Identification of topology errors in power system state estimation
,
1993
.
[6]
Sukhan Lee,et al.
A Gaussian potential function network with hierarchically self-organizing learning
,
1991,
Neural Networks.
[7]
Lamine Mili,et al.
A robust estimation method for topology error identification
,
1999
.