Partners Selection in Multi-Agent Systems by Using Linear and Non-linear Approaches

Abstract

Traditional negotiation approaches pay intensive attention to decision making models in order to reach the optimal agreements, while placing insufficient efforts on the problem of partner selection. In open and dynamic environments, when the number of potential partners is huge, it may be expensive or even impractical to perform complicated negotiations with all of its potential partners. In this paper, based on the proposed extended dual model, we propose both linear and non-linear approaches for partner selection in multi-agent systems. By employing these two approaches with the extended dual concern model, agents can adapt their individual behaviors for partners selection in negotiation. The proposed approaches have three merits, which are: (1) both agents' own benefits and their potential partners' benefits are considered during the partners selection process; (2) agents' preferences are employed by the proposed approaches which ensure the selection results to accord with agents' expectations; (3) the proposed approaches are sensitive to changes of the negotiation environment, so they can be adopted in open and dynamic negotiation environments. According to the case study in four scenarios, the selection results are reasonable and accord with agents' expectations.