Fuzzy inference systems approach for long term hydrothermal scheduling / Sistema de inferencia nebulosa ao planejamento da operação hidrotermica de medio prazo

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

The long term hydrothermal scheduling lies in the optimization of the water resource usage through the maximization of the hydroelectric production and the minimization of the thermal plants operation. Its goal is to assure an economic and reliable load supply throughout the study stages. This problem can be characterized by exhibiting a complex nature, since its characteristics define it as a large scale, dynamic, stochastic and nonlinear problem. Although many optimization approaches have already been proposed to answer the hydrothermal scheduling problem, until now, there is no unanimous approach that is able to cope efficiently with all the problem issues. Dynamic Programming, which is one of the most commonly used techniques to deal with this problem, is limited regarding its application on real systems since its computational requirements tend to be heavy. In this work we proposed an alternative approach to deal with the long term hydrothermal scheduling. The proposed technique is based on an Adaptive Neuro-Fuzzy Inference System working in parallel with a deterministic optimization model with perfect inflows forecasting. The optimal operation information is processed by the network that produces fuzzy rules describing the optimal decisions of the plant through the definition of the amount of discharge in each stage and depending on the chosen input variables. The performance of the proposed Neuro-Fuzzy approach was compared to other policies, including Deterministic Dynamic Programming, Stochastic Dynamic Programming and Open- Loop Feedback Control, by simulation using historical inflow records of five different Brazilian hydroelectric power plants. The results demonstrated that the Neuro-Fuzzy approach provided similar and competitive performance to less computationally efficient and commonly used policies

ASSUNTO(S)

optimization methods otimização matematica simulação (computadores) sistemas de energia eletrica power systems hydroelectric power generation fuzzy neural networks usinas hidroeletricas hydrothermal scheduling logica difusa

ACESSO AO ARTIGO

Documentos Relacionados