An agent-based evolutionary strategic negotiation for project dynamic scheduling

摘要

In this paper, an agent-based approach with a mutual influencing, many-issue, one-to-many-party, strategic negotiation model is proposed. The model concentrates on solving the dynamic scheduling problem of a distributed project for non-cooperative and self-interested participants. In this model, the self-interested activity agents possess various negotiation tactics and strategies formed by their respective owner's subjective preference, aim to find the contract of schedule adjustment mutually acceptable to respective participant's acquaintance while encountering conflicts over rescheduling settlement. In order to find fitting negotiation tactics and strategies that are optimally adapted for each activity agent, an evolutionary computation approach which encodes the parameters of tactics and strategies of an agent as genes in GAs is also addressed. In the final, a prototype system with a case of a distributed project for dynamic scheduling discussed in researches is simulated to validate the feasibility and applicability of the approach, and some characteristics and future works are also addressed.