A new realistic optimization-free economic load dispatch method based on maps gathered from sliced fuel-cost curves

Abstract

Many methods have been introduced as economic load dispatchers. However, all these methods depend on the optimization field, where most of these studies mainly focus on how to strengthen the optimization algorithms themselves. This means the mystery key that segregate between the good and bad ELD solvers is the optimization algorithm itself. However, there is a practical fact known in many real power stations that the generators set-points, assigned by automation centers, are actually in discrete form. Thus, the solutions presented in the literature are infeasible if they are seen from this practical point of view. One of the approaches is to use combinational optimization algorithms. Because most of ELD problems have limited number of generating units, so all the solutions can be extracted by slicing the fuel-cost function of each unit in order to have a full map of all feasible solutions. Based on that, the exact optimal solution as well as all other best solutions can be easily obtained by this method. The main challenge that could be faced with this optimization-free method is when the dimension of the given ELD problem is large, because the discrete search space exponentially increases as the number of variables or/and slicing resolution increases. This challenge can be avoided if that ELD problem is correctly formulated based on real configurations of electric power stations. Some numerical experiments are given to illustrate the mechanism of the proposed technique.