Narcís Nabona (Universitat Politècnica de Catalunya, Barcelona)

Long-term electric power planning through the Bloom and Gallant formulation. Models and optimization techniques
Monday 16 December 2002 at 15.30, JCMB 6310

Abstract

Bloom and Gallant have proposed an elegant model for finding the optimal thermal schedule subject to matching the load-duration curve and general linear constraints. Their method is based on a linear program with some linear equality constraints and many linear inequality constraints. There have been applications of this procedure to multi-interval problems using the active set method and the Dantzig-Wolfe column generation method, and through direct application of linear optimization packages using an available modeling language.

This presentation describes the model of long-term hydrothermal electric power planning adapted to use Bloom and Gallant's procedure, and puts forward a new quadratic formulation of the maximum profit problem of a generating company in a competitive market.

The implementation of the active set and the Dantzig-Wolfe column generation methods to solve these problems will then be discussed.

The computational experience reported includes the solutions of several long-term power planning problems of different sizes using a modeling language and available linear and quadratic programming solvers, and the comparison of the performance of the active set method and the Dantzig-Wolfe column generation procedure. Several remarks will be made on the effect of implementation issues in both procedures.

Seminars by year

Current 2019 2018 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996