A contribution to the midterm scheduling of large scale hydrothermal power systems

A concise mathematical model and nonlinear minimization algorithm were presented as a joint approach to set optimal midterm production targets for power systems with predominance of hydroelectric power generation. The algorithm uses the network structure of the problem and takes into account nondifferentiable points. The approach was successfully applied to the optimal midterm scheduling of a very large hydroelectric power system in southeastern Brazil. Further insight into the problem allowed the adoption of an optimization-simulation strategy which achieved a very good suboptimal solution with great savings in computer requirements. The main purpose of this research was to develop a working tool able to assist engineers in setting optimized midterm targets for the large power systems in Brazil, where more than 90% of the electric energy comes from hydraulic sources.>