|M.Sc Student||Mor Michal|
|Subject||Resource Allocation and Contingent Branching: An|
|Department||Department of Aerospace Engineering||Supervisors||PROF. Yoseph Ben-Asher|
|PROFESSOR EMERITUS Yakov Ben-Haim|
The first part of this research deals with resource allocation under unstructured uncertainty (unexpected or uncontrollable occurrences). A resource allocation problem is analyzed using the info-gap approach, chosen specifically because of its ability to contend with unstructured uncertainties. In this problem we consider the allocation of fuel to civilian aircraft in an international airport, to which the supply of jet fuel has been cut off. The cut-off resource, jet-fuel, acts as a constraint, and both the amount of jet-fuel needed and the duration of supply interruption are sources of uncertainty. The idea was to offer a way to deal non-probabilistically with problems that have traditionally been treated with probabilistic approaches.
In the second part, an extension to the information-gap approach is presented, analyzed, and applied to an example. This extension is designed to deal with strategic decision-making where there is contingent branching in the flow-chart. Contingent branching has been studied before, but not from an information-gap approach point of view. In addition, two sets of theoretical questions tied to contingent branching are examined. The first set deals with the notion of flexibility: the first proposition in this set states that flexibility does not invariably enhance robustness, and the second states that the flexibility does not invariably enhance opportunity. The second set of questions deals with the notion of redundancy: the first proposition in this set states that the redundancy invariably reduces robustness, and the second states that redundancy invariably reduces opportunity.
This thesis opens the way to further research into the application of the information-gap approach to more resource allocation problems, and into further analysis of the contingent branching extension in terms of more complicated node structures and application to more types of problems.