An improved unscented Kalman filter applied into gps positioning system

In the GPS positioning system, the traditional unscented Kalman filter has a large amount of calculation, so it is unable to meet the real-time requirement. And if the dynamic model is affected by abnormal disturbances error, the accuracy and stability of unscented Kalman filter is influenced, too. To solve those problems, we use a minimum skewness monomorphic sampling strategy to reduce the amount of calculation of unscented Kalman filter and improve the accuracy of unscented Kalman filter. It can reduce the impact of the abnormal disturbances error on the accuracy of unscented Kalman filter by adjusting the process noise adaptively. So we propose an improved unscented Kalman filter algorithm for GPS positioning system. Simulation results show that the accuracy of improved unscented Kalman filter algorithm is better than traditional unscented Kalman filter algorithm. © 2015 ICIC International.