D. Peharda, M. Delimar, and Z. Hebel (Croatia)
power system, state estimation, neural network.
Power system control and analysis rely on power system state estimation for the information on the current state of the system. Conventional state estimation methods are computationally intensive and may give unpredictable results when redundancy is low and gross errors are present. This paper explores the potential of the application of neural networks in the area of static state estimation. Several types of neural networks were tried out. Best results were obtained with a multilayer perceptron with linear input and output layer. Multilayer perceptron development and results are discussed in detail.
Important Links:
Go Back
