多Agent协商竞争模型的研究及仿真

Study and Simulation of Multi-Agent Negotiation Competition Model

在利用多Agent构建人工系统的过程中,Agent的利益偏好不同及资源有限等因素会引起冲突,而多Agent之间有效的协商竞争能够减少冲突,提高整体性能。为此,提出一种新的多Agent协商竞争模型,该模型依托协商信息,提高竞争对手策略预测准确率,同时采用概率最大化利益行为选择策略,从而缩短协商竞争时间。

In the process of building manual system relying on multi-Agent,the factors such as the difference of Agent＇s interests,the limited resources and so on will inevitably lead to conflict.While effective consultation and competition between multi-Agent helps to reduce conflict and improve overall performance.It proposes a new multi-Agent negotiation competition model,which relies on negotiation information to improve the forecast accuracy of opponent＇s competing strategies,depends on probability to selection strategies to maximize the interest and shortens negotiation time effectively.