CKF And Its Application On Spacecraft Autonomous Navigation

Spacecraft autonomous navigation is the development direction of space technology.Considering the launch,in-orbit operation and orbit maneuver phases of spacecraft,whole segment autonomous navigation algorithms are studied,taking celestial navigation and GNSS as measurement.This paper analyzes the navigation satellite system observability,on this basis,to deal with problems,like nonlinearity,abnormal external force and uncertainty of measurement and control,etc.,navigation filters are proposed,taking the CKF as the basic filtering algorithm,which achieve the whole process adaptive navigation.The main work includes the following three parts:1,The observability of satellite autonomous navigation system is studied.The state model of the navigation system is established according to the orbit dynamics,and the observation model based on Sun-Earth-Moon astronomical information is given;the observability theory of the continuous linear system and discrete linear system is introduced,and the observable conditions of continuous nonlinear systems and discrete nonlinear systems are obtained according to the observability theory of nonlinear systems;the observability of the autonomous navigation system is analyzed and the basic condition of non observability is put forward.2,An autonomous navigation algorithm based on robust CKF filter is studied.A spherical-radial cubature rule-the core of CKF is clarified based on the theory of Bayesian filtering,then,he basic process of CKF navigation filter is derived,the estimation accuracy of CKF filter and UKF filter are compared,which proved that the precision of CKF is better than that of UKF for high dimensional nonlinear systems,meanwhile,the basic process of CKF navigation filter is derived.In view of the uncertainty system,and The robust CKF is proposed adopting the M-estimation and the chi-square testing,which improves the robustness of the filter.3,The CKF navigation filtering algorithm based on dynamic model compensation is proposed.It’s hard to achieve high precision spacecraft navigation at launch,orbit maneuver and in-orbit operation phases since the forces spacecraft suffered change a lot and noise of measurement is too large.Thus,the statistical weighted fusion scheme is proposed.Measurement data of each channel are weighted adaptively which can restrain the error of observers and enhance the reliability of the system.Meanwhile,according to the characteristics of dynamic model error,a dynamic model compensation algorithm based on CKF filter is constructed,and the model error is estimated by expanding the state and adding the model error to the state equation,simulation shows that the proposed algorithm can effectively realize overall process and the whole segment autonomous navigation.