An improved tracking model based on self-adaptive Kalman filter

The self-adaptive Kalman filter algorithm can solve the problem of filtering divergence and tracking in conventional Kalman filter algorithm, so it is widely used in GPS data processing of vehicles and ships. But in practical application, the tracking of adaptive algorithm is still insufficient. Therefore, this paper proposes an improved tracking model based on adaptive Kalman filter. In this paper, by adding velocity and acceleration constraints to the conventional Kalman equation, we find that the equation only changes in the one-step prediction equation. By adding the changing parameters to the self-adaptive parameters in self-adaptive Kalman filter algorithm, it can effectively enhance the tracking ability of the adaptive algorithm to the velocity and acceleration changes. At last, the improved model is improved by multi-model and VSIFT. The simulation results show that the improved method can effectively enhance the model tracking performance.