Scheduling as a fuzzy multiple criteria optimization problem

Real-world scheduling is decision making under vague constraints of diierent importance, often using uncertain data, where compromises between antagonistic criteria are allowed. We explain in theory and by detailed examples a new combination of fuzzy set based constraints and repair based heuristics that help to model these scheduling problems. We simplify the mathematics needed for a method of eliciting the criteria's importances from human experts. We introduce a new consistency test for connguration changes. This test also helps to evaluate the sensitivity to connguration changes. We describe the implementation of these concepts in our fuzzy constraint library ConFLIP++ and in our heuristic repair library D ejj aVu. Finally, we present results from scheduling a continuous caster unit in a steel plant.