首页> 外文会议>IEEE International Conference on Tools with Artificial Intelligence >Strategy Inference in Multi-Agent Multi-Team Scenarios
【24h】

Strategy Inference in Multi-Agent Multi-Team Scenarios

机译:多代理多团队方案中的策略推断

获取原文

摘要

Creating simulations for multi-agent multi-team interactions is a daunting task. It is non-trivial to compose a situation where each individual agent maintains their own 'personality' while still following the assigned policy dictated by a team's central command. Further, the complexity is inflated by ensuring that each of these agent policies is coordinated into a cohesive team strategy. Finally, peaking the complexity, is evaluating the performance of the team's strategy against other teams' strategies in real-time. This is the work of this paper, proposing SIMAMT, the simulation space for multi-agent multi-team engagements, and testing it. We will first cover the system and how well it models the virtual environment for strategic interaction. Second, we will deliver results from a practical test of strategy inference within such an environment using the SIE (Strategy Inference Engine).
机译:为多主体多团队互动创建模拟是一项艰巨的任务。构成这样一种情况是很简单的,在这种情况下,每个特工都保持自己的“个性”,同时仍然遵循团队中央司令部规定的分配策略。此外,通过确保将这些座席策略中的每一个协调成一个紧密的团队战略,可以提高复杂性。最后,要达到最高的复杂性,就是要实时评估团队策略相对于其他团队策略的绩效。这是本文的工作,提出了SIMAMT(用于多主体多团队参与的模拟空间)并进行了测试。我们将首先介绍该系统及其为战略交互建模虚拟环境的能力。其次,我们将使用SIE(策略推断引擎)在这种环境下提供对策略推断的实际测试结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号