Study for outliers based on Kalman filtering

In Kalman filtering applications the observation values including outliers are important effect on optimal filtering. The outliers affect the orthogonal property of Kalman filter innovation sequence, filtering accuracy and make estimation inaccurate. In this paper, the modifying orthogonal property of innovation sequence based on Robust Kalman filtering is presented, which can keep orthogonal properties of innovation sequence. Simulation results show that the modified algorithms are effectively resistant to outliers in sampling data.