基于MINE-FNA组合算法的防空体系指标网构建方法?

防空作战体系的武器装备组成结构和对抗信息交互具有网络化特征，其作战效能是通过对抗过程中各组分系统间复杂的网络化交互涌现出来的，因此描述体系作战效能的指标也应由各组分系统效能指标相互关联涌现出来，而传统防空体系指标选取方法无法准确描述防空体系内部复杂的关联关系和揭示体系能力的复杂涌现生成机理。因此将无参数探索最大交互信息算法（ MINE）和网络统计算法（ FNA）组合构建了MINE?FNA组合算法模型，对基于武器装备体系仿真试验床的防空体系仿真的全样本实验数据进行深度挖掘，构建了“底层基础指标→均一化指标→体系级指标”的三层防空体系指标网，为研究防空体系内部复杂关联关系和分析防空体系能力的生成机理提供了有效手段。

Indices Network Construction for Air Defense System of Systems Based on MINE-FNA Assembled Algorithm

The physical composition structure and information exchange in Air defense operation system of systems ( ADOSoS) exhibit network feature. Operational effectiveness emerge from the complex network interaction between the inter?component system in ADOSoS, therefore the capability index also emerge from the inter?component system index. Traditional capability index selection methods could not accurately reveal the internal relationship and the emergence mechanism of the capability of the ADOSoS. In this paper, we built the assembled machine learning model ( MINE?FNA) , combined the Maxi?mal Information based Nonparametric Exploration with Feasible Network Analysis algorithm, to farm the multidimensional and holographic operation data deeply. According to the model, we build the three?tier Indices Network:Underlying index→Homogenization index→ADOSoS index, which can provide an effective means to analyze correlation and emerging mecha?nisms in ADOSoS.