Windowing-based random weighting fitting of systematic model errors for dynamic vehicle navigation

The Kalman filter is a commonly used computational method for dynamic vehicle navigation and positioning. However, it requires kinematic and observation models not contain any systematic error; otherwise, the resultant navigation solution will be biased or even divergent. In order to overcome this limitation, this paper presents a new windowing-based random weighting method to fit the systematic errors of kinematic and observation models within a moving time window for dynamic vehicle navigation. This method compensates the systematic model errors by correcting observation residual vector and state noise vector during the filtering process. Random weighting theories are established to fit the systematic model errors and the covariance matrices of observation vector and predicted state vector within a moving time window. Experiments and comparison analysis with the existing methods demonstrate that the proposed method can effectively resist the disturbances on system state estimation due to the systematic errors of kinematic and observation models, thus significantly improving the accuracy of dynamic vehicle navigation