动态系统的抗差Kalman滤波及其影响函数

Robust Kalman Filter and Error Influence Functions for Dynamic System

离散历元动态观测量及其相应的动态模型可能存在异常,若在数据处理中不考虑对这些异常的特别处理,则动态模型参数估值及其所提供的动态信息将极不可靠。对于GPS动态定位数据的处理,按抗差估计原理,构造了状态向量和观测值对模糊度的影响函数,并由此建立了动态GPS定位的抗差Kalman滤波解法。实际计算验证了该方法的实用性和可靠性。

With the effect on the Kalman filter caused by the abnormal of the observations being considered in GPS kinematic positioning, the robust Kalman filter is constructed, and the error influence functions corresponding to the estimation of ambiguity are derived. The theory of GPS kinematic positioning with robust Kalman filter is formed.