Aplicação de metaheurísticas ao despacho econômico com restrições

Abstract

Economic Dispatch (ED) is classified as a problem whose objective is to minimize the costs associated with the allocation of the output power of generating units, subject to the physical and operational constraints of the system. Among the main constraints are the so-called valvepoint effect (VPE), transmission system losses, ramp rate limits, and prohibited operating zones. Two IEEE test systems, comprising 6 and 15 thermal generating units, were employed for the development of the proposed methodology, which uses two metaheuristics: the Sine- Cosine Algorithm (SCA) and the Equilibrium Optimizer (EO) separately. The constraints are handled through penalization of the objective function, and the stopping criterion is the maximum number of iterations. The results were compared with those from other studies in literature that used metaheuristics to solve the ED problem in the same test systems. Both metaheuristics met the total load demand and the operational constraints. In the case of the smaller system, the SCA achieved the lowest total minimum cost among the analyzed studies, while the EO provided a lower average total cost over the course of 50 runs. In the larger system, the EO stood out in both aspects: it presented the lowest minimum and average costs among the studies analyzed. The standard deviations obtained in both systems, for both algorithms, showed relatively high values compared to the literature. This characteristic is attributed to the stochastic nature of the two methods, which was intentionally maintained to broaden the exploratory reach throughout the executions. Furthermore, the convergence achieved for the analyzed systems was highly satisfactory, confirming the strong global and local search capabilities, as well as the high efficiency of both applied heuristics, which prove to be very promising for solving the economic dispatch problem in static systems of realistic size.