| With the rapid development of GNSS systems,Satellite navigation system can be as the main means of navigation of civil aviation.Aviation navigation must be high precision,high security and high reliability,which puts forward high demands on the integrity of the navigation system.Receiver autonomous integrity monitoring(RAIM)method is a technology that not dependent on external system and can provide users with navigation integrity monitoring in time.At present the study focuses on the methods that use statistical test theory and the least-square to meet the integrity requirements of system.In order to solve the interference of gross error to the least squares and the statistical test values,RAIM algorithm is constructed in this paper by using M estimation method of robust estimation.The following research has carried out in this paper:1)Fault detection and elimination algorithm is designed in this paper according to the robust properties of M-estimation.The actual observation data is used for gross error detection experiments,fault detection and elimination experiments are done using several common M-estimation methods,such as L1-norm-estimation,the sum maximal likelihood estimation,Huber method,IGG ? method,to treat single gross error,double or three gross errors of different sizes.M-estimation performance and usability are analyzed using the results of experiments.Conclusion can be draw after verification that the performance of M-estimation to treat gross error is significantly better than least-square.Especially the L1-norm-estimation method and the sum maximal likelihood estimation method can obviously improve the success rate of gross error detection and elimination.2)New algorithms are designed in this paper to calculate horizontal protect level(HPL)and vertical protect level(VPL)to forecast and evaluate the quality of estimation results of L1-norm-estimation.A horizontal and vertical protection level algorithm based on minimum detectable bias(MDB)of L1-norm-estimation is proposed.The results obtained by the SLOPE method and the MDB method are compared and analyzed.This paper proposes a multi-direction maximum slope(MMS)algorithm based on Bias Integrity Threat(BIT)and the position covariance matrix and pseudo range covariance matrix of the L1-norm-estimationto estimate the protection level,and the results obtained by the multiple hypothesis Solution Separation(MHSS)method and the MMS method are compared and analyzed.Compared with 95% vertical positioning error(VPE),the VPL based on the L1 norm estimation MDB method and the MMS method can be used to fully protect the VPE.The HPL and VPL results of the MDB method have some degree of improvement compared with the traditional SLOPE method,and the HPL and VPL of the MMS method also have a certain degree of improvement compared with the MHSS method.3)In this paper,the basic performance of BDS14 is analyzed in the service area,and the effects of failures of different types of satellite on the performance of RAIM in BDS system are analyzed.The simulation of GPS,BDS,GLONASS and Galileo constellations is done for test.The MMS algorithm which is proposed in this paper is used to calculate the VPL of a single system and multi-system combination of BDS,GPS,GLONASS and Galileo in the case of single gross error and two or three gross errors.The VPL is compared with the VPA requirements of different navigation phases and the global availability of navigation system is analyzed.In the case analysis,increasing number of satellites can reduce VPL value and increase the availability of the RAIM system when the accuracy of satellites is close;but the join of satellites with low accuracy will increase VPL value and reduce the system availability.When the number of faults increases,the availability of RAIM will be reduced,but the satellite geometric structure and the standard deviation of the satellite orbit and clock error with respect to the number of gross errors have a greater impact on the vertical availability of RAIM. |
PDF Full Text Request