Authors: L. Ramesh; Niladri Chakraborty
Addresses: Dr. M.G.R Educational and Research Institute, Maduravoyal, Chennai, India ' Jadavpur University, Kolkotta, India
Abstract: Online state estimation is becoming one of the key functions in deregulated distribution control centres. The bus voltages in the system are estimated from the measured values of feeder real and reactive power flows, etc. To reduce the metre cost, few actual measurements are carried out on the system through meter placement and the rest of the measurement need is satisfied by using pseudo-measurements, where estimation accuracy is often affected. To improve the estimation accuracy, the swarm tuned artificial neural network approach for bus voltage estimation in radial distribution systems is presented, which does not require pseudo-measurements. The back propagation neural network is trained by minimising the error function in a search space based on weights, tuned using particle swarm. The training data for the ANN is collected from real time monitoring of the test system through PSCAD simulation. The algorithm is tested to IEEE and TNEB network systems.
Keywords: artificial intelligence; power distribution systems; online state estimation; performance evaluation; bus voltages; artificial neural networks; ANNs; particle swarm optimisation; PSO; simulation.
International Journal of Artificial Intelligence and Soft Computing, 2016 Vol.5 No.3, pp.209 - 225
Accepted: 06 Sep 2015
Published online: 22 Aug 2016 *