欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
多Agent合作追捕是多Agent系统研究的经典问题，在机器人等领域具有重要的应用前景。文章提出了面向任务的自利Agent联盟生成算法，该算法能同时处理多个不同类型的逃跑Agent的任务分配问题；因为追捕Agent和逃跑Agent速度相等，追捕联盟成员位置的分布十分关键，所以提出了基于贡献度的联盟成员选择策略；同时为了较好地体现追捕Agent的自利性，定义了需求度作为自利性的度量，解决了冲突协商且有利于资源的优化配置。通过与经典拍卖算法的比较表明，该算法显著提高了追捕成功率。
英文摘要：
Multi-Agent cooperative pursuit is a classic problem of multi-Agent systems, which has important applications in the field of robotics. The alliance generation algorithm of task-oriented self-interested Agent is proposed. The algorithm can handle the task allocation problem of many different types of escape Agents. For the pursuit Agent and escape Agent have equal speed, the position distri- bution of alliance members is critical. Therefore, an alliance member selection strategy based on con- tribution degree is put forward. In order to better reflect the self-interest of pursuit Agent, need de- gree is defined as a measure of self-interest, which can deal with the conflict negotiation and promote the most optimum allocation of resources. Compared with the classic auction algorithm, the proposed algorithm significantly improves the success rate of the pursuit.
同期刊论文项目
同项目期刊论文
Hand Gesture Modeling and Recognition for Human and Robot Interactive Assembly Using Hidden Markov M
期刊信息
位置：