论文部分内容阅读
针对多agent系统的具体网络结构,提出一种分布式联盟形成算法.首先,每个agent向自己邻居任务中“信价比”最高的任务提出加入申请,形成潜在联盟;当联盟资源大于对应任务的资源需求时,使相应agent退出当前联盟,转而选择“信价比”次高的任务,从而提高联盟形成的可能性.仿真实验结果表明,与其他方法相比,所提出的算法能大幅提高系统效率和收益.
Aiming at the specific network structure of multi-agent system, a distributed coalition formation algorithm is proposed.Firstly, each agent puts forward an application to the task with the highest “letter-to-price ratio” in its neighbor task to form a potential alliance. When the coalition resource is greater than Corresponding to the resource demand of the task, the corresponding agent is withdrawn from the current coalition, and the task of “higher letter-to-price ratio” is selected to increase the possibility of formation of the coalition. The simulation results show that compared with other methods, The algorithm can dramatically improve system efficiency and profitability.