Fast static available transfer capability determination using radial basis function neural network

In a competitive electricity market, available transfer capability information is required by market participants as well as the system operator for secure operation of the power system. The on-line updating of available transfer capability information requires a fast and accurate method for its determination. This paper proposes a radial basis function neural network based method for available transfer capability estimation in an electricity market having bilateral as well as multilateral transactions. Euclidean distance based clustering technique has been employed to select the number of hidden radial basis function units and unit centres for the radial basis function neural network. In order to reduce the number of inputs and the size of the neural network, a feature selection has been performed using two different methods based on Euclidean distance based clustering and random forest technique and the performance of the radial basis function neural network, trained with features selected using these two methods, has been compared. The effectiveness of the proposed method has been tested on 39-bus New England system and a practical 246-bus Indian system. (C) 2010 Elsevier B.V. All rights reserved.