How can a decision maker take into account the evolution of his preferences over time?

TL;DR

This paper introduces dynamic flow balances to help decision makers adapt to evolving preferences in industrial processes, exemplified by Munster cheese production, enabling smoother adjustments and forecasting trends.

Contribution

It proposes a novel technique called dynamic flow balances for modeling preference evolution and adapting industrial decisions accordingly.

Findings

Dynamic flow balances forecast production trends.

The method helps avoid oscillations and consumer annoyance.

Potential for multi-criteria predictive control.

Abstract

The ranking of possible alternatives during the design or operation of an industrial process leads to the definition of a solution representing the best compromise between several contradictory objectives. The choice of solution is a decision that can be modeled using knowledge of the corresponding preferences. The decision and preference model used is that of flow balances. Munster cheese is used as an example. These are highly distinctive cheeses that are not suitable for uninformed consumers. If the decision-maker wishes to widen his consumer panel, he must modify the production line settings. In the event of a change in preferences due to an external influence, the decision-maker must adapt, but not too abruptly, to avoid both production line oscillations and consumer annoyance. To solve this problem, we propose a new technique: dynamic flow balances. This makes it possible to forecast production trends. Prospects are also proposed, such as the creation of a multi-criteria predictive control.