Research on the selection of the basic parameters in deep-space navigation system

The classic six orbit elements and the position-velocity vectors are two modes to describe the states in deep-space navigation system. So, how to choose the proper mode is an essential problem. Based on the nonlinear character and the space structure character, we researched the state equations, the measurement equations and the applicability of the nonlinear Kalman filter algorithm. Then we imported the nonlinearity concept to analyze the observability and explain the phenomenon that the observability is different using the two modes. At last, the simulation of the navigation system in cruise phase was performed to confirm the above theory. The results show that, using orbit elements, combined with Kalman filter based on nominal trajectory expansion, we can get higher navigation accuracy. So the deep-space navigation system, especially the cruise phase should use the orbit elements to describe the states. However, we must judge the nonlinearity of the system before calculating the observability with the linearization method.