Design of an application for the solution of economic dispatch problem of power system considering risk and using soft computing methods

Abstract

The solution of Economic Dispatch problem has been based on deterministic approach, where the input parameters are known with complete certainty. However, a lot of inaccuracies and uncertainties are present in the solution. Thus, involving risk that represents these uncertainties is necessary for finding an optimal solution. This paper provides a design of a new technique for the solution of Economic Dispatch problem considering risk. Recently, emphasis is placed on risk based solutions rather than probabilistic solution [2]. The design proposed in this paper gives an insight into generating risk based solution of Economic Dispatch problem using concept of variation, Fuzzy sets and Artificial Neural Network (ANN). Since there are uncertainties involved in the solution, a coefficient of variation is used for each of the generator. Fuzzy sets are used for finding membership values. These membership values are given as inputs to ANN. Using these membership values as inputs, ANN computes a penalty term representing risk. The Lagrange multiplier method is employed for the solution of Economic Dispatch problem which uses the penalty term. This paper provides description of the various strategies used in the solution considering risk.