Abstract:In this paper, the Helmert variance component estimation based on Kalman filter is introduced to adjust the weights of observations in combined GPS/BDS precise point positioning. Meanwhile, the Kalman filter prediction value is taken as a quasi observation; it can be treated using variance component estimation with observations for the difference in unit variance between prediction value weight ratio and observations weight ratio. The experimental results show that this method can reasonably determine various observations weight ratio of the current epoch. Moreover, it can effectively adjust the value of the observation and the value of Kalman filter. Thus, the positioning accuracy and reliability of GPS/BDS combined PPP are obviously improved.
LONG Xin,ZHANG Xianyun,DENG Xiaodong et al. Helmert Variance Component Estimation Considering Quasi Observation of Kalman Filter and Its Application in Combined GPS/BDS PPP[J]. jgg, 2018, 38(11): 1149-1152.