Metaheurísticas bio-inspiradas aplicadas em problemas de despacho econômico de carga

Abstract

The growing consumption of electricity is forming a highly competitive and vibrant consumer market, changing many aspects of electricity generation, transmission and distribution. The search for sufficient energy resources to meet the load demand, the high cost of electricity generation and the growing concern to mitigate environmental impacts, directly influence the economic load dispatch (ED), which is essential for the generating units of electricity reach the optimal minimum cost of operation. The problem of ED, firstly, is to minimize the cost of generating electricity and operating the generating units, meeting the total load demand and the inherent restrictions of the system, thus configuring an optimization problem. In this context, the objective of this paper is to solve four problems ED considering generation systems with three, six, fifteen and twenty thermoelectric units. In all cases, transmission line losses and maximum and minimum generation limits for each generating unit are considered. For this, heuristics were used: particle swarm optimization (PSO), firefly algorithm (FA), cuckoo search algorithm (CS) and gray wolf optimization (GWO). All methods showed total cost and transmission line losses compatible with those found in the literature. The results obtained were also compared with the solution achieved by the Lagrange Multiplier Method (LMM). Thus, the CS stood out for the excellent total cost solutions obtained and a low standard deviation index, good convergence and efficiency in determining the exact solution or very close that verified through the LMM. For the system with twenty machines, the solution obtained by GWO was better than the achieved by PSO, FA and CS. This solution, when compared to those found in the specialized literature, is also better. The GWO did not obtain the minimum cost in any of the analyzed cases, however, it obtained very close to the optimal solutions, with relatively low computational cost, in relation to the other analyzed methods. The PSO presented minimum or near optimal cost in most cases, with relatively high computational cost. The FA presented solutions close to the optimal and low computational cost. The CS presented high computational cost, however, obtained excellent solutions. Thus, it was possible to conclude that the CS methods and GWO performed better the task of optimizing the economical load dispatch for the analyzed problems.