An efficient graph partition method for fault section estimation in large-scale power network

In order to make fault section estimation (FSE) in large scale power networks using a distributed artificial intelligence approach, we have to develop an efficient way to partition the large-scale power network into the desired number of connected sub-networks such that each sub-network should have balanced working burden in performing FSE. In this paper, a new efficient multiple-way graph partition method is suggested for the partition task. The method consists of three basic steps. The first step is to form the weighted depth-first-search tree of the power network. The second step is to further partition the network into connected balanced sub-networks. The last step is an iterative process, which tries to minimize the number of the frontier nodes of the sub-networks in order to reduce the required interaction of the adjacent sub-networks. The proposed graph partition approach has been implemented with applications of sparse storage technique. It is further tested in the IEEE 14-bus, 30-bus and 118-bus systems respectively. Computer simulation results show that the proposed multiple-way graph partition approach is suitable for FSE in large-scale power networks and is compared favorably with other graph partition methods suggested in references.