The stochastic short-term hydrothermal scheduling Problem

Abstract

In this document we develop a non linear stochastic integer model formulation for the unit commitment problem of thermal and hydro units to management demand and optimal short-term operation of a hydrothermal electric facility. We consider a power generation system comprising thermal and hydro units and the problem concerns the scheduling of operation levels for all power units and considering the hydro constrains, such that the operation costs over the time horizon are minimal. The concept of reliability functions is introduced to ensure the meet demand with certain probability. Inflows to reservoirs, cost coefficients and spillage are considered random. We use the Monte Carlo sampling to optimize the instances required each period. We report the practical and theoretical results.