一个多Agent学习模型

A MODEL OF MULTIAGENT SYSTEM LEARNING

多Agent学习问题近年来引起人工智能领域的广泛注意 ,本文给出了一个基于遗传算法的多Agent学习模式 ,学习过程分为四个部分 :任务的分解、子任务的分配、多个子任务并行求解和解答的集成。其中遗传算法主要应用在子任务的分配和多个子任务并行求解中。

Recently the problem of multi agent learning has increased the attention of AI area more widely.This paper introduces the common definition of multi agent learning.Then a learning mode based on genetic algorithm is discussed deliberately.The learning mode is composed of four parts:the deposition of problem,the distribution of child problem,the multi child problem solution,and the last one is the sum of the solution.The genetic algorithm is mostly applied to the distribution of child problem and the multi child problem solution.