A (Soft) Robustness for Possibilistic Optimization Problems

This paper discusses a linear programming problem and a general combinatorial optimization problem with uncertain parameters, whose unknown distributions are modeled by fuzzy intervals. The fuzzy intervals have possibilistic interpretation. Some criteria for choosing robust solutions to the problems under consideration, resulting from the use of possibilistic decision theory, are proposed. It is shown that the fuzzy problems constructed are computationally tractable if their deterministic counterparts are polynomially solvable. The algorithms for finding the robust solutions are provided. Some computational experiments are performed.