Research In Application Of GPS Kinematics Navigation Filtering Algorithms

Global Position System (GPS) has recently gained extensive developments in the aerial, marine and land vehicle navigation fields. But in regard to GPS's positioning accuracy, there are so many error sources found from GPS satellite to GPS signal. One of the methods that can reduce the positioning error of GPS is GPS kinematics filtering, which has lower cost of hardware than DGPS without DGPS base station and digital communicational arrangement. Even if the signal is received by differential positioning of GPS, it can be processed by kinematics filtering again, which still can improve positioning accuracy.First of all, error sources of GPS positioning are analyzed. The discussion focuses on two important objective factors that influence positioning accuracy: localization signal and the receiver. Furthermore, GPS signal is simulated.In addition, to the nonlinear problem about state estimation of GPS dynamic navigation system, based on the shortcomings in the EKF about dealing with nonlinear filtering, a kind of improved derivative-free kalman filter is created, which uses square-root of covariance matrix (it can be obtained by Cholesky decomposition) instead of covariance matrix to iterative computation, then formulas of state and random noise covariance matrix's Cholesky decomposition update are classified. The improvements are made to avoid derivative calculating and non-positive definiteness of covariance matrix's square root, accelerate the speed of filtering, and inhibit divergence. The improved algorithm is applied to state estimation for GPS/DR integrated navigation system. Simulation results show that the improved algorithm has higher superiority than EKF and UKF.Finally, based on the vehicle simulation experiment, the schema about dynamic test of GPS is discussed, using a couple of GPS receiver platforms, which are different in accuracy with one order of magnitude, and practicality of the new kalman filter is verified, considering filtering comparison between DGPS and GPS of the uniform motion vehicle. Results of the simulation experiment and test show that the improved derivative-free kalman filter has not only better accuracy and stability, but higher practicality than traditional kalman filter.