Techniques of the Optimization Based in Artificial Intelligence Applied to Hydrothermal Power Systems Operation Planning

The Hydrothermal Power Systems Operation Planning (HPSOP) aims to determine a strategy for each generation plant at each time interval in order to minimize the expected value of operating costs in the planning horizon. This is a challenge for managers of the Energy Sector, because of the stochastic nature of the problem which is coupled in time (dynamic) and in space (interconnected), large, not separable, non-convex and the target function is non-linear. Therefore, the application of classical techniques presents several limitations. In order to overcome those limitations, improvement of traditional methods, or the development of alternative heuristics is a vital step in the operation of HPSOP. This paper presents an application of two of this new heuristics of Artificial Intelligence: Genetic Algorithms and Ant Colony Optimization. Those techniques were applied to a test system with data from Brazilian power plants. The results showed a good performance when compared with traditional optimization techniques already used in HPSOP. It is noteworthy that in the current study, the applications of traditional optimization techniques and Artificial Intelligence have made use of real characteristics of plant operation without the need to simplify the original formulation.