Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Tractable Reserve Scheduling Formulations for Power Systems with Uncertain Generation


V. Rostampour

Master Thesis, HS12 (10261) Politecnico di Milano

The increased penetration of renewable energy sources (in particular, wind generation) into the electricity grid highlights the necessity of constructing stochastic variants of the so-called unit commitment (deciding which generators to switch on at what time) and reserve scheduling problems. In this thesis, we formulate an optimization problem for a DC power flow based model of power network and propose a technique based on randomized optimization to solve both the unit commitment and the reserve scheduling problems in a unified framework while providing probabilistic performance guarantees. This technique allows us to handle integer variables, an essential step for addressing the unit commitment problem. The resulting performance enhancement in terms of the total energy procurement cost observed in simulations was quite substantial. We then introduce affine feedback policies for reserve scheduling to assess whether more sophisticated "recourse" strategies are able to provide a further improvement in energy costs. Finally, we extend our power network models to more accurate models based on AC power flow and propose a novel relaxation method to obtain a convex (albeit very complex) optimization problem. Empirical evidence shows that the proposed relaxation technique tends to a feasible solution for the exact problem. The theoretical developments in all cases above were validated on realistic benchmark problems of power networks and a discussion on the tractability of the resulting optimization problems is presented.


Type of Publication:

(12)Diploma/Master Thesis

K. Margellos

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2013:IFA_4372
Permanent link