Many optimization problems are represented with algebraic equations, using continuous and discrete variables, giving rise to Mixed Integer (Linear or Nonlinear) Programs (MILP/ MINLP). An alternative way to represent this type of problems is Generalized Disjunctive Programming (GDP) proposed by Raman and Grossmann[1] that involves not only algebraic equations, but also disjunctions and logic propositions. This higher level representation, which is of special importance in engineering, allows us to exploit the logic structure of the problem to obtain better formulations and improved solution methods.
展开▼