A genetic algorithm approach to optimize the operation planning of hydrothermal system scheduling

This article investigates the application of Genetic Algorithms to optimize large, non-linear complex systems, particularly the optimization of the operation planning of Hydrothermal System Scheduling Several of the current works to solve this kind of problems are based on non-linear programming. However, this approach present a number of deficiencies, like difficult convergence, oversimplification of the original problem or difficulties related to the approximation to the objective function. Aiming to find more efficient solutions for this class of problems, this paper proposes and investigates the use of a genetic approach. The authors believe that, due to their characteristics, like simplicity, parallelism and generality, Genetic Algorithms can provide an effective solution to the optimization of the operation planning of Hydrothermal System Scheduling. For such, this paper shows the application of this technique for the optimization of the operation planning for a system composed by three hydroelectric plants located irt the Brazilian Southeast System.