Stochastic streamflow model for hydroelectric systems using clustering techniques

In this paper, clustering techniques were applied in the monthly streamflow generation model developed for the Brazilian hydroelectric system to alleviate the computational effort in the mid-term operation planning model. The work is organized in three parts. The first part describes briefly the model for calculating mid-term optimal operating strategies in a multi-reservoir hydroelectric system. The second part presents the monthly streamflow model based on autoregressive modeling of periodic hydrologic series. The last part describes multivariate techniques that are used to group objects based on their intrinsic properties. A case study is used to illustrate the performance of this approach.