ARTIFICIAL NEURAL-NETWORK APPROACH TO NETWORK RECONFIGURATION FOR LOSS MINIMIZATION IN DISTRIBUTION NETWORKS

Network reconfiguration of distribution systems is an operation in con figuration management that determines the switching operations for a m inimum loss condition. An artificial neural network (ANN)-based networ k reconfiguration method is developed to solve the network reconfigura tion problem to reduce the real power loss in distribution networks. T raining-sets for the ANN are generated by varying the constant P-Q loa d models and carrying out the off-line network reconfiguration simulat ions. The developed ANN model is based on the multilayer perceptron ne twork and training is done by the back propagation algorithm. The trai ned ANN models determine the optimum switching status of the dynamic s witches along the feeders of the network, which thereby reduce real po wer loss by network reconfiguration. The proposed ANN method is applie d to the 16-bus test system. Test results indicate that the developed ANN models can provide accurate and fast prediction of optimum switchi ng decisions for minimum loss configuration. The proposed ANN method i s compared with Kim's method [IEEE Transactions on Power Delivery 8, 1 356-1366 (1993)] and a comparative study is presented. The proposed me thod can achieve minimum loss configuration with drastic reductions in the number of ANNs and less computational time. (C) 1998 Elsevier Sci ence Ltd. All rights reserved.