Study On Techniques Of GNSS Integrity Simulation And Monitoring

The integrity is a key performance indicator in the global satellite navigation system(GNSS), the simulation and processing technology is very important to ensure the construction, reliable operation of the GNSS. Especially for the Beidou satellite navigation system(BDS), which started late and the technology requirements are rigorous. The the state of the system is determined by the integrity simulation system, which can greatly reduce system development risk. At the same time, the performance of the integrity monitoring system is further enhanced in the treatment of multiple faults, small pseudorange deviation and other faults, which can effectively promote the BDS in more areas. Therefore, it has important engineering value and practical significance to research the integrity simulation and processing technology. Under this background, this paper has done the following research work:(1) For the GNSS integrity simulation system, using piecewise high order polynomial model to calculate the changed pseudorange contained the integrity faulty, there is contradiction between pseudorange calculation accuracy and real-time, and the accuracy of commonly pseudorange high order derivative calculation method is poor in the high dynamic scene, the satellite orbit calculation method based on the Window Lagrange Interpolation(WLI) and pseudorange derivative calculation method based on three order Hermite interpolation are presented to solve these problems, and with the combination of both, a new high precision real-time pseudo distance and its derivative calculation method is got. The simulation results show that, when the derivation step is 0.001 s and the time interval is 1s, compared with the three order spline interpolation method, the standard deviation of pseudorange error in proposed method is decreased from 0.369 mm to 0.172 mm, the calculation accuracy is improved by about 1 times, but the calculation quantity is just 10%.(2) In the GNSS integrity simulation system built in the real equipment, the traditional dynamic delay generation method using the simulation to generate the navigation signals existed some problems, such as consistency with the time domain and related domain delay measurement, consideration of delay control complexity and resource consumption, a Variable Fractional Delay(VFD) filter design method based on the maximum group delay error and a dynamic delay generation method based on the VFD filter and are proposed respectively. The design examples show that, the maximum group delay error in the proposed design method is decreased from 3.596×10-3 to 9.287×10-4, the delay accuracy is improved by about 74.17% compared with the iteration weighting least square menthod.At the same time, under the condition that the accuracy of the distance, speed, acceleration and jerk are respectively 1 cm, 1 mm/s, 1 cm/s2 and 1 cm/s3, the maximum delay error is 0.8 mm, the maximum code doppler error is 0.02 m Hz, for the proposed dynamic time delaygeneration method.(3) The traditional multiple faults receiver autonomous integrity monitoring(RAIM) algorithm has large quantity for the calculation, A new multiple faults RAIM algorithm based on signal-to-noise anomaly detection is proposed to monitor the multiple satellite failure presented in user segment. In this algorithm, the fault satellites are detected by calculating the detection statistics of signal to noise ratio anomaly in real-time. The process for multiple faults RAIM algorithm contain the optimal selection of satellites, positioning solution estimation and receiver clock error inspection, thus significantly reducing the computational burden of fault detection and isolation. The analytical results in real data show that, in 3 failure satellites simulation conditions, the position error in the proposed method is decreased from 14.6m to 3.5 m, the position accuracy is improved by 76% compared to the traditional RAIM method. Further the statistic results are consistent with theoretical values by analyzing the probability characteristics of signal-to-noise linear model and anomaly detection statistic.(4) The existing RAIM algorithm cannot effectively detect tiny and slow continuous pseudo-range bias, with the non coherent accumulation increased normalized equivalent pseudorange deviation, the model of fault detection was established, the parity vector method based on non coherent accumulation and the robust Extend Kalman Filter(EKF) method based on prediction residual sliding accumulation are proposed respectively. The simulation results show that, for the improved parity vector method, it is insensitive to the the deviation symbol. Under the same detection performance, the detectable minimum pseudo distance deviation in 10 accumulation epochs is reduced from 2.4m of the traditional method to the 0.78 m. And for the improved resistance EKF method, when the pseudo distance deviation change rate is 0.01m/s, the length of the sliding window of 10 conditions, the positioning error in the proposed method is reduced for 1.056 m to 0.801 m, the positioning precision is improved 24.1% compared with the robust EKF method.