基于极限学习机的目标智能威胁感知技术

空中目标威胁感知是智能空战的关键技术之一。针对空中目标威胁评估问题，提出一种基于极限学习机的目标威胁智能感知策略。该方法首先提取目标威胁评估的高权重态势因子，并采用隶属度函数进行归一化数值解译；然后，借助专家知识采用极限学习机理论对威胁感知的输入输出数据进行建模，构建智能感知推理模型；最后，建立基于极限学习机的目标威胁智能感知流程。仿真结果表明，该算法具有较高的威胁感知精度以及较好的算法实时性。

Intelligent Threat Perception of Aerial Target Based on Extreme Learning Machine

Threat perception of aerial target is one of the key technologies in intelligent air combat. Strategy based extreme learning machine (ELM) is proposed to cope with the threat assessment problem. In the proposed method, the high weighting situation factors impacting on target threat are selected firstly, and then normalized and quantized via membership functions. The mapping relationship between the situation factors and target threat is modeled based on ELM. Finally, the threat perception process of target is established. The simulation results show the proposed algorithm can obtain high perception accuracy and realize online learning.