一种改进的系统偏差估计算法

One Modified Algorithms to Estimate System Errors

雷达组网系统首先要解决误差配准问题,用以准确地估计和消除系统偏差。实时质量控制RTQC算法是一种经典的配准算法,但是该算法提出时并没有考虑随机量测误差对配准性能的影响,现对这一问题进行了分析,在其基础上构造了一种关于量测和系统偏差的线性模型。然后根据这一线性模型提出了一种改进的基于广义最小二乘的误差配准算法,并对算法的性能进行了比较和分析。仿真结果表明,本文提出的算法能有效地进行误差配准。

Registration problem is a prerequisite process for radar networking system to estimate and correct systematic errors accurately. The Real-Time Quality Control （RTQC） algorithm is a classical approach for registration. In this paper, we get the precision of the RTQC algorithm by considering the effect of the random measurement noise, and then construct the linear model between the measurement and the system errors. According to the linear model, two modified registration algorithm are proposed, which are based on least square estimation and general least square estimation. The performance of the three algorithms is analyzed and compared. Simulation data approve that these two algorithms can alignment the system biases effectively.