一种改进的雷达组网误差配准算法

Modified Registration Algorithm for Radar Networking

雷达组网系统首先要解决误差配准问题,来准确地估计和消除系统误差。对一种经典的误差配准技术广义最小二乘估计GLSE配准算法进行了研究,讨论了系统偏差增大时,配准精度退化的原因,并针对这一问题提出了一种改进的误差配准算法(MGLSE算法)。通过循环配准,MGLSE算法不但可以对真实的系统偏差进行逼近,而且当GLSE算法失效时,可以自动退出配准循环,放弃配准。仿真结果表明,系统偏差较大时MGLSE算法优于GLSE算法和其他传统的配准算法,并具有配准效果自动检验功能,防止配准后系统偏差反而增大。

problem is a prerequisite process for radar networking system to estimate and correct systematic errors accurately. Generalized linear least-squares estimation (GLSE) algorithm is a classical registration algorithm, but the registration precision will deteriorate while the system errors increase. The reason is discussed, and then a modified GLSE registration algorithm is presented, which can approach the real system errors by means of the cycle of GLSE registration. If the GLSE algorithm becomes invalid, the cycle of registration will terminate, and the registration will be abolished. Simulation results show that the MGLSE algorithm is superior to the GLSE algorithm and some other classical registration algorithms, while the system errors are not sufficiently small. Simulation results also show that the MGLSE algorithm can examine the registration effect automatically to avoid the deterioration of the system errors after registration.