On the mixing set with a knapsack constraint (Q291057)

The paper investigates a substructure that arises in mixed integer programming reformulations of chance-constrained programs with stochastic right hand sides over a finite discrete distribution. This structure is called the mixing set with knapsack constraint and is denoted by \(Q\). The authors study the structure of the convex hull of \(Q\), generalizing earlier results for the case of equal probabilities, where the knapsack constraint reduces to a cardinality constraint. They characterize valid inequalities that do not come from the knapsack polytope. Then they propose a relaxation scheme based on a polyhedral relaxation of the knapsack polytope and derive facet-defining inequalities for the convex hull of \(Q\). The paper concludes with numerical experiments.