| 潜艇CGF规避反潜航空兵行为建模 |
| |
| 引用本文： | 徐旺,曹志敏,韩瑜.潜艇CGF规避反潜航空兵行为建模[J].指挥控制与仿真,2012,34(6):103-109. |
| |
| 作者姓名： | 徐旺 曹志敏 韩瑜 |
| |
| 作者单位： | 江苏自动化研究所,江苏 连云港 222006 |
| |
| 摘 要： | 对战场态势的不完全观测是潜艇作战的重要特点,充分利用不完全观测信息,提高潜艇对抗决策的合理性是设计潜艇CGF的核心问题。针对潜艇作战特点,提出了潜艇自防御行为模型框架,并利用多Agent不确定场景建模技术,建立了基于POMDP的规避反潜航空兵行为模型,最后进行了仿真验证。结果表明,该模型具有较强的可行性。
|
| 关 键 词： | 计算机生成兵力 POMDP 强化学习 行为建模 |
| 收稿时间： | 1/1/1900 12:00:00 AM |
POMDP for submarine CGF behavior modeling during the confrontation with an ASW Patrol Aircraft |
| |
| xuwang,cao zhi min and han yu.POMDP for submarine CGF behavior modeling during the confrontation with an ASW Patrol Aircraft[J].Command Control & Simulation,2012,34(6):103-109. |
| |
| Authors: | xuwang cao zhi min and han yu |
| |
| Institution: | Jiangsu Automation Research Institute,jiangsu Automation Research Institute,jiangsu Automation Research Institute |
| |
| Abstract: | Making full use of the Partial Observed information to keep Submarine from the threat is the foundation of the intelligent CGF design, since the observation of submarine is always incomplete and unreal. In this paper, we will cast the defense system of submarine CGF as a POMDP, which is a popular method of stochastic modeling in multi-agent systems. Finally, we check our model by a confrontation with an ASW Patrol Aircraft, and the results also show its great feasibility. |
| |
| Keywords: | Computer Generated Forces POMDP Reinforce Learning Behavior Modeling |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《指挥控制与仿真》浏览原始摘要信息 |
| 点击此处可从《指挥控制与仿真》下载免费的PDF全文 |
|
