A hybrid algorithm for mid-long term hydrothermal generation scheduling

The paper presents a deterministic model, which takes into account the final water storage and the effects of water head, for optimal mid-long term generation scheduling of a hydrothermal system. Considering its large scale mixed integer nonlinear structure, a hybrid algorithm based on Genetic Algorithm and decomposition algorithm is developed to solve it. The model is divided into thermal power dispatch and hydro power dispatch subproblems respectively by using reservoir volumes as genetic variables which consider the complex relationship between reservoirs. These two subproblems are independently solved according to their own characteristics. The simulation results reveal that comparing with GA the hybrid algorithm is an effective method with shorter computation time and higher precision.