粗集支持向量机的战斗机空战效能智能评估

摘要

根据现代空战特点,选取了战斗机的空战效能评估指标集,并采用粗糙集理论对指标体系进行约简,提取对空战效能影响起关键作用的特征参数,消除冗余信息,减少了支持向量的维效.支持向量机( SVM)具有结构简单、全局最优、泛化能力强的优点.根据所提取的特征参数,文中提出采用回归型支持向量机(SVR)建立空战效能智能评估模型,并通过实例与指数法和BP神经网络法计算结果进行了比较,验证了该模型的可行性和有效性.%According to the characteristic of modern aerial warfare. the index set of aerial warfare efficiency evaluation is selected in this paper. Reduction is performed on index systems based on rough set theory to extract characteristic parameters which affect aerial warfare efficiency crucially, which is to remove redundant information and reduce the dimension of support vector. Support vector machine (SVM) has the advantages of simple structure, global optimum and high generalization ability. With the characteristic parameters, intelligent evaluation model for aerial warfare efficiency of fighter-plane is establish by using Support Vector Regression (SVR), and we compare the SVR with index method and BP network method by a case study, which verified the feasibility and validity of the model.