Kalman滤波抗野值方法研究

在Kalman滤波应用过程中 ,观测值中的野值是影响滤波效果的重要因素。当观测中含有野值时 ,破坏了Kalman滤波新息的原有特性 ,从而造成估计不准 ,滤波精度下降。本文提出了修正Kalman滤波新息的方法 ,使修正后的Kalman滤波新息能够保持修正前的新息特性。仿真结果表明 ,本文提出的方法可有效地抑制观测中的野值对系统滤波的不利影响

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 properties of Kalman filter innovation, filtering accuracy and make estimation to inaccurate. In this paper the modifying innovation based on Robust Kalman filtering is presented, which can keep initial properties. Simulation results show that the modified innovation filtering algorithms efficiently resistant to outliers in sampling data.