Intelligent approach for efficient operation of electrical distribution automation systems

Distribution systems play a vital role in providing an efficient service in terms of power quality, reliability, and economy. Distribution network reconfiguration can be used for planning as well as real time control. The paper presents an efficient approach for network reconfiguration based on artificial neural networks. A package, called "DISTFLOW", is developed adopting the proposed technique. The off-line simulation results and daily load curve data are used for training the neural network. Further, the distribution system operation is optimized by selecting an optimum compensation level computed by genetic algorithms (GA). The proposed integrated approach is applied to a practical 140 bus system in the Surathkal city subdivision of the power utility Mangalore Electricity Supply Company (MESCOM).