Rational Decision-making under Uncertainty in Transport: Minimising Maximum Regret

摘要

Transport decision-makers generally choose between projects in the context of uncertainty. Projects are characterised by multiple attributes, so multi-criteria decision-making tools are required. However, uncertainty relates both to the acceptability of attribute values and to the actual values of the attributes themselves. The Bellman-Zadeh fuzzy decision principle caters for uncertainty in the set of acceptable attribute values by treating this as a fuzzy set. Kikuchi and Chakroborty introduce possibility theory to cater for both uncertainty in the attribute values and uncertainty in the set of acceptable attribute values. However, this approach requires the specification of a possibility distribution for an attribute value and appears to be limited to decisions based on a single attribute or objective. This paper replaces possibility distributions by intervals for attribute values, bases decisions on the concept of regret, and extends the decision framework to encompass multiple objectives by synthesis with the Bellman-Zadeh fuzzy decision principle. Two illustrative numerical examples are presented.