论文部分内容阅读
对 GPS定位中的 SA误差作了系统辨识。表明 :SA误差可用时变 AR( P)模型表述 ,AR( P)的时变参数可由 RLS算法结合 F判据建立的 AR( P)模型辨识机获得。用该方法获得的模型对 SA误差进行预报 ,在剔除野值后精度在 1 0 m之内 ,还提出一种方法 ,可得到具有 SA预报能力的 GPS KALMAN滤波器。
The SA error in GPS positioning was systematically identified. It shows that SA error can be described by time-varying AR (P) model and time-varying parameter of AR (P) can be obtained by RLS algorithm combined with AR (P) model recognition machine established by F criterion. The model obtained by this method is used to predict the SA error. After removing the outliers, the accuracy is within 10 m. A GPS KALMAN filter with SA prediction capability is also proposed.