Complexity Analysis of Neural Approaches Used to Solve Economic Dispatch Problems in Power Systems

Several neural approaches have been proposed in the literature to solve problems of power dispatch involved with electrical energy systems. More specifically, taking into account just the power economic dispatch, the Hopfield network is the neural model more used in this application type, and most of the neural approaches developed are variations of the Hopfield network. The proposal of this paper is in investigating the computation structure carried out in each neural approach, whose final objective is the accomplishment of complexity analysis relating to these different neural models. As result of this investigation, it is possible to define some criteria that indicate which are the more appropriate neural approaches in relation to the particular characteristics of the economic dispatch problems.