Multi Dimension of Coarse to Fine Search Method Development for Solving Economic Dispatch

Abstract

Economic dispatch problem has grown along with the development of electric power business, for example in a competitive electric power business that offers electrical energy in the form of the step function, non-differentiable function. This is not a continuous function so there is no guarantee that those methods can execute the optimization problem well, especially the Lagrange and Direct methods. There are the non-differentiable functions within the optimization will become a challenge that should be solved. This paper proposes Coarse to Fine Search method development to solve the problem. The Coarse to Fine Search is able to work for differentiable or non-differentiable functions, but is only limited maximum three dimensions. The development is done to multi dimension so that it can solve the economic dispatch problem. We named it Multi Dimension of Coarse to Fine Search. The simulation results of eight power plants show the developed method can work well, it is always convergent and fast with the execution time of 2.63 - 38.30 seconds for 25 - 200 population and 50 - 200 delta search.