一种基于航迹特征的无人机与飞鸟目标雷达识别方法

A Radar Recognition Method for UAV and Flying Bird Targets Based on Track Characteristics

对于现代雷达探测系统而言,无人机与飞鸟同属于具有“低慢小”特征的一类典型目标,而面对比较复杂的作战环境,其对功能的要求已经不仅局限于对两者目标实现稳定探测跟踪,如何有效区分两者类型并完成识别更是当下急迫且重要的难题。常规方法是从目标的微动特征差异进行区分,但由于两者回波微弱,很难通过时频分析方法提取目标特征。针对该问题,从航迹特征出发,提出一种无人机与飞鸟目标雷达识别方法。首先对比两者目标在运动轨迹上的差异性,进行特征分析,提出时间相关的航向震荡频率与速度震荡频率特征量描述方法,并在离线状态下,利用实测雷达系统记录的航迹数据,提取两者的有效特征量;然后利用支持向量机算法对样本进行训练,并在获得最优模型参数后,通过测试样本进行测试,测试分类结果显示准确识别率能够达87%;最后在线状态下跟飞实验,其结果既表明该方法的正确性,也体现了在工程实现角度上的轻量性、实用性、适用性,具有较高价值。

For modern radar detection systems,unmanned aerial vehicles(UAVs) and birds belong to a typical type of targets with "low,slow,and small" characteristics In complex combat environments,the functional requirements of radar detection systems are not only limited to achieving stable detection and tracking of the two targets How to effectively distinguish the two types and complete the recognition is an urgent and important challenge at present Conventionally,the targets are distinguished from the differences in their micro-motion characteristics However,it is difficult to extract the target features through time-frequency analysis methods since the amplitudes of the two echoes are very weak In order to solve this problem,in this paper,a radar recognition method for UAVs and flying bird targets is proposed based on track characteristics First,the differences in the motion trajectories of the two targets are compared,then a feature analysis is conducted,and a time-dependent description method for heading the oscillation frequency and velocity oscillation frequency feature quantities is proposed In the offline state,the effective feature quantities of the two targets are extracted from the track data recorded by the actual radar system Then,the samples are trained by the support vector machine algorithm After the optimal model parameters are obtained,tests are carried out The test classification results show that the accuracy of the identification can reach 87% Finally,flight tests in the online state are conducted The obtained results not only indicate the correctness of the method,but also reflect its lightweight,practicality,and applicability in the perspective of engineering implementation,which has high value.