Pareto-optimal equilibrium points in non-cooperative multi-objective optimization problems

摘要

In this paper, we consider a class of multi-objective optimization (MOP) problems where the objective holders are independent humans or human-based entities. These problems are indeed game problems, which we call noncooperative multi-objective optimization problems (NC-MOP). We discuss that for such problems, the ParetoOptimal (PO) solutions are not necessarily valid as they primarily require Nash equilibrium (NE) solutions. Instead, we suggest that a new solution concept of the Pareto-optimal Equilibrium (POE) point could be adopted. Such a solution is, in particular, important in engineering design and articulation of new rules and protocols among independent entities. This paper reviews all relevant works that approach the POE concept and investigates the interplay between game problems and multi-objective optimization problems. We present illustrative examples to deepen our understanding of where a POE solution is achievable, as this is not always the case.