Optimized Design of the Navigation Filter in Ultra-Tight GPS/INS Integration

Under extremely high dynamic conditions, GPS receiver cannot exactly track carrier frequency, even if it is equipped with baseband signal pre-filter. Since the tracking errors of different channels are not independent any longer after passing through the traditional navigation filter, the ultra-tight GPS/INS integration may crash. To solve this problem, the pseudorange rate tracking errors were modeled and treated as the state variables of the navigation filter. However, the computational complexity of the filter with expanded dimension is too high. So on the premise of keeping navigation accuracy, dimension reduction was carried out through the use of inter-satellite single-difference and two-stage Kalman filtering algorithm. Finally, a semi-physical simulation, based on the gyro and accelerometer measurements telemetered in a rocket missile test, validates the effectiveness of the proposed methods.