基于遗传算法的SVM自适应干扰样式选择

Selection of SVM Adaptive Interference Mode Based on Genetic Algorithm

针对目前自适应干扰决策中存在的干扰样式匹配准确率不高、干扰实时性低等问题,在传统遗传算法的基础上,提出一种改进遗传算法(improved genetic algorithm,IGA)优化的支持向量机(support vector machine,SVM)算法,用于干扰样式自适应选择。采用IGA对SVM的惩罚参数和核函数参数进行优化,强化模型的学习能力和泛化能力,提高算法在干扰决策中的实时性和准确率。分别从干扰决策的准确率和实时性两个方面,与传统基于网格搜索(grid search,GS)法优化的SVM模型进行对比。仿真结果表明,IGA-SVM模型在进行自适应干扰样式选择时,干扰决策的实时性和干扰样式匹配准确率相对于传统网格搜索法有一定提高。

In order to solve the problem of low real-time and low matching accuracy of interference mode in adaptive interference decision making,this paper presents a SVM based on IGA for interference mode adaptive selection.The penalty parameters and kernel function parameters of SVM were optimized by IGA to enhance the learning ability and generalization ability of the model and improve the real-time and accuracy of interference decision making.IGA-SVM is compared with SVM based on GS method in terms of the accuracy and real-time of interference decision.The simulation results show that,in adaptive interference decision making,the real-time performance and interference mode matching accuracy of IGA-SVM are improved compared to traditional GS-SVM.