The field of stochastic programming provides a framework for developing models, methods and theory for optimization under uncertainty and facilitates many challenging applications.
The first part of the thesis studies the inclusion of risk measures in the traditional expectation-based models and contains results on the structural properties and algorithmic development for the so-called mean-risk models.
Motivated by the recent liberalization of the electricity markets and the resulting increase in market price uncertainty, the second part of the thesis presents a number of new stochastic programming applications to power systems. Of particular importance are the day-ahead market bidding and the subsequent re-scheduling of a hydro-power producer who is further subjected to inflow uncertainty. As an important part of modeling, special emphasis has been placed on generating price and inflow scenarios.