基于增程雷达一维像特征的高轨卫星识别方法

高轨卫星距离地球数万公里,雷达为了实现有效跟踪,通常使用增程技术。远距离使位置测量误差大到千米量级,增程导致数据率低到每帧以分钟计,雷达散射截面积(RCS)细节信息丢失,这给利用目标轨道和RCS进行目标分辨、识别带来困难。幸运的是高轨卫星目标相对观测视线位置和姿态稳定,天然消除了宽带一维像使用中方位敏感性的问题。本文通过一维像的尺寸、波形熵等特征,通过最小二乘支撑向量机,设计了高轨卫星目标的模板匹配算法,利用实测数据进行了识别验证,结果表明了方法的有效性。

High-Orbit Satellite Recognition Method Based on HRRP Characteristics of Extended Range Radar

Since high-orbit satellites are tens of thousands of kilometers away from the earth, in order to achieve effective tracking, radar usually uses the extended range technology. The long distance makes the error of position measurement be the order of kilometers, the extended range causes the data rate as low as in minutes per frame, and thus the details of the radar cross session (RCS) are lost. These make it difficult to distinguish and recognize targets by using the target orbit and RCS. Fortunately, the relative line-of-sight position and attitude of high-orbit satellite targets are stable, which naturally eliminates the orientation sensitivity problem in high resolution range profile (HRRP) applications. In this paper, a template matching algorithm for high-orbit satellite targets is designed based on the characteristics such as HRRP size and waveform entropy with the least squares support vector machine. The algorithm is identified and verified by measured data. The results show that the proposed method is effective.