摘要
提出了一个分布式解决无人驾驶坦克自主作战决策的框架,将自主作战、动态任务规划和任务策略的智能评估等思想引入无人驾驶坦克自主作战决策系统。包括以下两个重要内容:主体个体的能力和职责;信息和知识如何在主体之间传播使得他们能准确地完成作战任务。文中重点是无人驾驶坦克自主作战决策系统中主体之间动态知识的交换。主体确保自己的行为是一种动态的经验操作和直接与其他主体交流事实、需求和规则的结合。在各个主体中,知识可以作为事实、规则、需求的结合体来传播交换。
Presents an uninfluenced combat decision system framework to support distributed problem solving in the uninhabited tank domain. The important ideas, such as uninfluenced combat, dynamic programming assignment and intelligent evaluating assignment, were advanced to the multi-agent system for uninfluenced combat decision in uninhabited tank. This paper addresses two important issues:the capacities and responsibility of individual agents;How the information and knowledge transfer should take place among agents to allow them to respond successfully to the battle assignment. The focus of this paper is dynamic knowledge exchange among uninfluenced combat decision system agents. A coordinator agent together with a decision enabling warehouse acting as a dynamic blackboard plus direct intercommunication among the agents enable facts, commands, and rules to be transferred between uninfluenced combat decision system agents. Knowledge can be exchanged among the agents by using a combination of facts, rules and commands transfers.
出处
《微机发展》
2004年第6期4-7,共4页
Microcomputer Development
基金
国家自然科学基金资助项目(70272002)
关键词
无人驾驶坦克
自主作战决策系统
分布式解决问题
多主体系统
uninhabited tank
uninfluenced combat decision system
distributed problem solving
multi-agent system