Mixed integer programming formulation techniques and applications to Unit Commitment problem

Abstract

Unit Commitment (UC) problem with non-linear functions and probabilistic constraints are difficult to solve by standard optimization methods. This paper provides an introduction to mixed integer programming problem formulation techniques and its applications to UC problem. Use of binary variables to convert nonlinear functions and probabilistic constraints to a Mixed Integer linear programming (MILP) form are presented. This conversion helps in solving the UC problem by using commercially available MILP solvers.