基于多Agent建模的弹道导弹多特征综合识别框架

摘要

由于弹道导弹目标特征具有积累时间各异、分类方法繁多、存在干扰等难点，需要系统具备智能、自适应和协作的能力来完成目标的综合识别。在传统综合识别的基础上，提出了一种基于多Agent建模的弹道导弹多特征综合识别框架，构建了Agent的推理规则库和内部状态库，描述了Agent的处理流程与交互关系。基于Swarm平台仿真验证表明，该框架能有效进行多特征综合识别，给出正确的识别结果。%It is difficult for radar to implement ballistic missile identification successfully, and it needs to integrate the capabilities of intelligence, self-adaption and cooperation for the identification system. On the basis of traditional identification methods, a multi-feature fusion identification framework based on multi-agent modeling is presented. The reference rule base and internal state base of the agents are established. The process procedure and interactive relationship of agents are given. The simulation based on Swarm platform shows that the multi-feature fusion identification framework can implement fusion identification effectively with correct recognition result.