基于递推增广最小二乘的雷达输出数据滤波方法

Method of Radar Output Data Filtering Based on the Recursive Augmented Least Square

针对雷达-光电经纬仪组网测量时雷达实时输出引导数据随机误差大的问题,提出采用递推增广最小二乘拟合算法进行实时滤波,建立了实时滤波模型,给出了基于DSP的雷达数据实时滤波总体框架,进行了软硬件实现,验证表明,新算法及DSP处理实现了雷达数据实时轨迹滤波、预测及输出,引导数据随机误差减小1个数量级,能够满足测量设备引导需求,实现对快速目标的平稳跟踪。

In the network measuring used photoelectric theodolite and radar,the real-time data of radar included lager error,this paper built algorithm based on recursive extended least squares fitting,and built the radar data filtering model, and designed the framework, and the hardware and sofware were implemented. Research result showed that adopt DSP curve filter realize data smoothness and extrapolate,and reduced random error 1 orders,which meets the guide requirement,and then achieves stable tracking of the fast projectile.