Optimization of the Hydrothermal Power Systems Operation Planning Based on Artificial Intelligence Techniques

Hydrothermal Power Systems Operation Planning (HPSOP) aims to define the generation strategy for each individual power plant which represents the minimal overall operational cost along the planning period. This is a challenge for the energy sector managers, because it is a large stochastic optimization problem, coupled in time (dynamic) and in space (interconnected). The objective function is nonlinear, not convex and inseparable. Classical techniques present difficulties to solve the HPSOP problem due to its complexity. Thus, improvements in the traditional techniques of nonlinear optimization or the development of alternative ones are essential to the HPSOP. This paper presents two artificial intelligence techniques: Genetic Algorithms and Ant Colony Optimization Algorithms. Both techniques were applied to a test scenario with two hydroelectric power plants from the Brazilian power system and they have demonstrated good results and performance when compared with the optimization techniques traditionally used for HPSOP. It is valuable to note that the optimization based on artificial intelligence techniques makes use of the actual operational characteristics of the power plants without the simplifications required by the traditional techniques.