基于异方差伪线性模型的纯方位目标要素估计

摘要

针对伪线性跟踪估计器假定误差项之间不相关且方差相等,导致利用最小二乘方法得到的目标要素的估计不是最优无偏估计的问题,提出了基于异方差伪线性模型的加权目标运动要素估计方法.该方法考虑了经典的伪线性模型误差项中随时间变化的异方差部分,首先通过方位信息估计出瞬时距离与初始距离比,得到伪线性模型误差项中的异方差部分的估计,然后利用加权最小二乘方法,得到了理论上目标要素的最优无偏估计.数值仿真和实验数据验证表明,该方法优于经典纯方位伪线性化方法.%The estimation of target elements using least square method are not optimal unbiased estimates as the pseudo linear target tracking filter based on non-correlation errors of model and equal variance.In order to solve the question,a heteroscedastic pseudo-linear model was established by using new parameter variables,based on the model,a weighted method for target motion analysis was presented.The heteroscedastic part of error of the pseudo-linear model was considered.Firstly,the ratio of instantaneous distance and initial distance were esti-mated by using bearings and the estimate of the heteroscedastic part of model error was obtained,then,By using the weighted least squared method,the optimal unbiased estimates of target elements were obtained in theory. Numerical simulation results showed that the new method was better than the classical pseudo-linear methodt.