SINS/GPS Integrated Navigation System Using an Improved Particle Filter based on State Reconstruction

Abstract

There exist nonlinear models in the integrated navigation system of strapdown inertial navigation system (SINS) and GPS. So it is appropriate to use particle filters to estimate the states. This paper focuses on the nonlinear problems when there exists large initial azimuth error in the SINS errors. In this paper, particles are driven to the regions of high probability by applying the error correction technique of Dynamic Matrix Control (DMC) to general particle filters and propose an improved particle filter. The proposed particle filter is then applied to the high-dimensional state model of SINS/GPS integrated navigation system. The simulation results show that the new algorithm doesn't need accurate error models of inertial measurement unit (IMU) but can still perform well and achieve more accurate estimates than unscented Kalman filter (KF).