FUZZY LEARNING DECOMPOSITION FOR THE SCHEDULING OF HYDROELECTRIC POWER-SYSTEMS

paper presents a nonlinear multivariable fitting model to decompose th e optimal policies obtained by dynamic programming of a unique aggrega ted reservoir. The nonlinear functions are generated using radial basi s functions (RBF) neural networks. In this method the potential energy of all the reservoirs in the hydropower system is added to form one e quivalent reservoir. The operating policy of the equivalent reservoir is determined by stochastic dynamic programming, and finally the opera ting rules of each reservoir are determined using RBF neural networks. To improve the multivariable representation of the data, a series of piecewise RBF neural networks is determined using clustering analysis. A fuzzy clustering approach is used to determine the RBF's parameters . This approach has the advantages of being fast and simple to impleme nt with well-established convergence properties. It also has a good re presentation of the covariance matrix, since all the data belong to al l the classes at the same time with different membership grades. A com parison with the back propagation learning and principal components te chniques is also reported for Quebec's La Grande River installations. As a result, the proposed approach gives satisfactory operating rules compared with principal component analysis, and the CPU time is reduce d by a factor of 15 to 20 compared with the back propagation technique .