A Novel MKPF and its Application on GPS/DR Integrated Navigation
Xiaoping Ma, Fangyuan Li · 2018
To improve the positioning accuracy of vehicles, combining GPS (Global Positioning System) and DR (Deadreckoning) for integrated navigation is a common way. For more popularized application, higher accuracy and more stable performance in vehicle positioning, more efficient data fusion methods are researched and applied. As the Common data fusion methods, Extended Kalman Filter(EKF) and Unscented Kalman Filter(UKF) could achieve better positioning results, but only under the condition of noise Gaussian distribution. And the EKF is only better in the small nonlinearity systems. Particle Filter (PF) is valued by scholars as its outstanding performance in nonlinear and non-Gaussian systems, and applied to GPS/DR integrated navigation. In order to improve the estimation performance of Particle Filters in GPS/DR integrated navigation, a novel MKPF method named STMKPF was designed and implemented, which combing Mixture Kalman Particle Filter (MKPF) with Strong Tracking Filter (STF).