Robust binary linear programming under implementation uncertainty

This paper studies binary linear programming problems in the presence of uncertainties that may cause solution values to change during implementation. type uncertainty, termed implementation is modeled explicitly affecting decision variables rather than model parameters. The nature invalidates use existing models for this uncertainty. robust solutions obtained are optimal a worst-case min-max objective and allow controlled degree infeasibility with respect associated deterministic problem. Structural properties used reformulate problem as mixed-integer program. conservatism by combining both constraint relaxation cardinality-constrained Solutions optimization under uncertainty consist set solutions; selection from possibly large formulated an over set. Results experimental study context knapsack suggest methodology yields perform well terms value feasibility. Furthermore, approach can identify possess desirable characteristics.