【摘要】 本文对方位测量野值存在的不同原因进行了分析 ,以提高目标参数估计性能的鲁棒性为目的 ,提出了 2种基于批处理技术的双基阵抗野值算法 ,分别是修正的最大似然估计法和特征值法 .并说明了特征值法和最大似然估计法一样 ,它也是一种无偏估计器 .蒙特卡洛仿真结果表明 ,运用这 2种算法得到的位置和速度均方根误差跟踪曲线显示了很好的野值抑制能力 ,并且修正的最大似然估计方法的抗野值能力要优于特征值方法 ,为工程应用提供了参考 .更多还原

【Abstract】 To improve the robustness of performance of target parameters estimatio n, different causes of bearings-only outliers’ existence are analyzed and two kinds of anti-outliers methods based on bistatic system are introduced. They ar e the modified maximum likelihood estimator and the so-called eigenvalue estima tor. It is shown that like the modified maximum likelihood estimator, the eig envalue estimator is also an unbiased one. Monte carlo simulation results show t hat both estimators can suppress measure outliers effectively in terms of positi on and velocity root mean square errors. Furthermore, the maximum likelihood estimator has better anti-outliers’ performance than the so-called eigenvalue estimato r, and it can provide references to engineering application.更多还原