摘要
为提高雷达有源压制干扰信号识别正确率,提出一种利用遗传算法优化神经网络的干扰信号分类识别方法。改进遗传算法,将种群中的个体按照适应度排序并等分为三段,在各段用不同比例选取相应数量的个体,同时用自适应的交叉率替换原固定取值,以加快算法搜寻速度;使用改进算法对前向反馈神经网络的权值和阈值进行优化;运用该网络进行雷达有源压制干扰信号的分类。针对射频噪声干扰、噪声调幅干扰、噪声调频干扰和噪声调相干扰4种典型压制干扰信号,仿真实验结果显示,所给方法可行有效。
In order to improve the correct recognition rate of radar active suppressive jamming signals,a new jamming signal classification and identification method based on genetic algorithm is proposed.The genetic algorithm is revised to speed up the search speed,the individuals in the population are sorted according to fitness and are equally divided into three sections. A corresponding number of individuals are selected with different proportions in each section,and the original fixed values are replaced by adaptive crossover rate.An optimization of weights and thresholds of back propagation neural network is implemented by using the revised algorithm,thus,this network can be used to classify the radar active suppressive jamming signals better.For the four typical jamming signals of RF noise jamming,noise AM jamming,noise FM jamming and noise PM jamming,the simulation results show that the proposed method is feasible and effective.
作者
杨洁
穆彦斌
程晓健
YANG Jie;MU Yanbin;CHENG Xiaojian(School of Communication and Information Engineering,Xi＇an University of Posts and Telecommunications,Xi＇an 710121,China)
出处
《西安邮电大学学报》
2018年第1期92-96,共5页
Journal of Xi’an University of Posts and Telecommunications
基金
国家自然科学基金资助项目(61402365)
陕西省科学技术研究与发展计划(工业科技攻关)资助项目(2013K06-33)
关键词
压制干扰信号
神经网络
遗传算法
适应度
suppressive jamming signal
neural network
genetic algorithm
fitness
