Hybrid Neural Network Models for Determination of Locational Marginal Price

Abstract

Locational Marginal Price (LMP) is one of the significant factors which plays a dominant role in the electricity market. Its determination using the conventional methods is highly laborious due to non-linearity. This paper presents three different Hybrid Neural Network models used for the determination of LMP. They are the genetic algorithm based neural network model, direct non-iterative state estimation based neural network model and a Radial Basis Function state estimation neural network model . A case study is made with the six bus test system. Test results are compared with those results obtained from a conventional Newton method and those of back propagation neural network model