Research On Vehicle Integrated Navigation Filtering Algorithms And Design Of GPS Module

Vehicle navigation is one of the most important technologies in intelligent transportation systems. Global Position System (GPS) is of superior long-term error performance, but poor short-term accuracy, while Dead Reckoning system (DR) has good position precision in short-term but poor in long-term. GPS/DR integrated navigation provides position data with high precision, frequency and reliability.On the base of analyzing traditional Extended Kalman Filtering (EKF), this paper puts forward a new method-interpolation filtering for integrated navigation systems. It adopts recursion structure of traditional EKF, and divided difference based on Stirling's interpolation formula substitutes for derivative calculation of nonlinear functions. Calculation is greatly simplified due to avoiding computing Jacobian matrix, and the new method improves model error caused by Taylor linearization in traditional EKF. DD1 and DD2 (first- and second-order divided difference) filtering methods correspond to first- and second-order function approximations based on interpolation formula. According to the state and observation models of GPS/DR integrated system, EKF, DD1 and DD2 filtering methods are compared to estimate state information of a given target. Simulation results indicate that DD1, DD2 filters have stronger tracking performance and larger application scope, and achieve higher numerical stability and accuracy than EKF.This paper proposes a design method of GPS module, which is a key device in vehicle integrated navigation systems. The module design uses GPS chipsets including LNA (Low Noise Amplifier), RF front-end and baseband processor and so on. Schematic circuit diagram and layout are provided to elaborate (1) LNA section design including Micro-strip for input matching and SAW filter; (2) RF front-end design concerning connections and structure of GPS front-end chip; (3) baseband design including GPSMODE configuration, active antenna supervision and extended memorizer circuits. Plate-making requirements are given as well. The design uses both GPS signal generator and real signal for module testing. Test results suggest that the GPS module has high sensitivity and good positioning performance. At last the module firmware configurations are described in brief.