Daniel Hamilton (University of Edinburgh)

Diet problems
Wednesday 27 January 2010 at 15.30, JCMB 6206

Abstract

Diet problems are optimisation problems where typically several raw materials are combined at minimum cost to make a product, subject to nutrient requirements.

The simplest type of diet problem is the Blending Problem, which was one of the first LPs solved. With the introduction of intermediate mixing bins the Pooling Problem is formed. We look at solving this NLP using SLP. When several factories are combined together we get a large NLP, that we call the Multifactory Problem. We solve this using decomposition by factory. We also look at using decomposition to find the global minimum to this highly nonconvex problem. A further type of diet problem that has integer constraints we call the Silo Problem. This is an MINLP, which we solve by decomposition into NLP and ILP.

Seminars by year

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