A Multi-Agent Alliance Forming Method Based on Balanced Alliance

Abstract

The invention discloses a multi-intelligent agent coalition formation method based on a balance coalition. According to the multi-intelligent agent coalition formation method, an intelligent agent task selective behavior model is established and a corresponding relation of an intelligent agent and a task is established, a step of calculating all potential coalitions firstly is reduced and the time of task allocation is saved. Furthermore, in combination with practical situations, each intelligent agent is firstly used as a main intelligent agent, only neighboring intelligent agents which can form a coalition with the main intelligent agent are considered, thus the problem that a result is affected because the coalition cannot be formed is avoided. In addition, according to the method, under the condition that the coalition capability cannot satisfy task requirements or the coalition capability is better than the task requirements, coalition members are recruited or reduced to realize the situation that the coalition capability can satisfy the task requirements, the redundancy of coalition resources is not generated, and the flexibility and the effectiveness of task allocation are improved.

Description

technical field [0001] The invention relates to the technical field of intelligent robots, in particular to a method for forming a multi-agent alliance based on a balanced alliance. Background technique [0002] In a multi-agent system, multi-agents can cooperate to complete tasks that a single agent cannot do, and ensure the smooth completion of tasks and improve the success rate of task completion. In the process of multi-agent cooperative task allocation, forming an alliance is an effective way to achieve multi-agent task allocation. The higher the balance coefficient of the alliance formed, the greater the fault tolerance rate of the alliance. [0003] The literature (Vig L, Adams J A. Multi-robot coalition formation [J]. Robotics, IEEE Transactions on, 2006, 22(4): 637-649.) gives a proof of a balanced coalition, and through all potential coalitions For the calculation of the balance coefficient, the alliance with the largest balance coefficient is selected as the fin...

AI Technical Summary

Problems solved by technology

However, in actual situations, some potential alliances cannot be formed due to abnormal and unexpected events such as communication congestion.

[0006] 3. In actual situations, most agents usually have multiple resources or abilities, but the resources or abilities of agents set in the literature are 1-dimensional numerical quantities, which do not conform to the actual situation

Images