Centralized Autonomous Orbit Determination Of BDS-3 Based On Adaptive Kalman Filter

The development of autonomous orbit determination technology based on inter satellite link system has become one of the hot issues in the research of satellite navigation system.In peacetime,autonomous orbit determination can improve the precision of satellite orbit determination by using inter satellite observation data under the condition of limited monitoring stations in China;In wartime,this technology can guarantee the wartime viability of navigation satellite system.However,when the navigation satellite is running in space,it will be affected by complex factors,which will lead to abnormal observation or carrier disturbance.It is very important to use adaptive Kalman filter to control the influence of observation value anomaly and dynamic model error in the course of autonomous orbit determination,which can improve the convergence speed of autonomous orbit determination filter and the accuracy and reliability of the solution results.This paper takes BDS-3 simulation constellation as the object,realizes the centralized autonomous orbit determination based on the adaptive Kalman filter,and discusses the accuracy of the autonomous orbit determination of the robust adaptive Kalman filter(AEKF)in the case of the observation value anomaly,the initial state information anomaly and the prior random information anomaly,with reference to the results of the centralized autonomous orbit determination based on the extended Kalman filter(EKF)The feasibility and effectiveness of robust adaptive Kalman filter for autonomous orbit determination of BDS are demonstrated.The specific work is as follows:1.Based on EKF algorithm,the influence of different ranging intervals,observation noises and residual system errors on autonomous orbit determination accuracy is discussed.(1)The influence of the distance interval between satellites on the orbit determination accuracy is studied.The experimental results show that the ur of autonomous orbit determination increases with the increase of ranging interval.When the inter satellite ranging interval increases from 5 minutes to 15 minutes,the ur increases by about 0.06 m on average.This shows that when the dynamic model of the system is not accurate,the prediction error of the dynamic model of the system will increase with the increase of the ranging interval.(2)Through the simulation experiment,the influence of observation noise and system error on the orbit determination accuracy is discussed.The results show that the random noise in the observations can be eliminated well in the filtering solution of autonomous orbit determination.The residual system error has a significant impact on the accuracy of autonomous orbit determination,and the fixed term error is the main part of the residual system error.There is no significant difference in orbit determination accuracy among the three kinds of satellites in BDS-3 constellation due to the change of observation noise and residual system error.2.Based on the EKF and AEKF algorithms,the effects of observation value anomaly,initial state information anomaly and prior random information anomaly on autonomous orbit determination accuracy are discussed.(1)Based on the EKF and AEKF algorithms,the influence of the initial state information anomaly on the convergence and accuracy of the autonomous orbit determination filter is studied.The experimental results show that the abnormal initial state information has a significant impact on the convergence and accuracy of the autonomous orbit determination filter.Among them,the influence time of position parameter anomaly on the accuracy of autonomous orbit determination results is longer;optical pressure parameter anomaly is the second;the influence of speed parameter anomaly on the accuracy of autonomous orbit determination results is larger,but the convergence speed is faster.The accuracy of the orbit determination results in AEKF scheme is better than that in EKF scheme.Among them,AEKF can effectively improve the influence of optical pressure parameter anomalies on autonomous orbit determination accuracy,which shows that AEKF has the ability to resist the influence of system dynamics model error.(2)The influence of the observation value anomaly on the orbit determination results of EKF scheme and AEKF scheme is studied.The orbit determination accuracy of AEKF scheme is better than that of EKF scheme,especially when dealing with observations with gross errors,AEKF scheme can significantly improve the accuracy of autonomous orbit determination results.(3)The influence of the prior standard deviation of observation noise and process noise on the orbit determination results of EKF and AEKF are studied.When the prior standard deviation configuration of process noise is abnormal,the ure results of AEKF scheme and EKF scheme diverge with time,but the result of AEKF scheme is better than that of EKF scheme;however,the change of prior standard deviation configuration of observation noise has no significant impact on the orbit determination results of AEKF scheme and EKF scheme.