Research On Deep Space Autonomous Navigation Scheme To Small Celestial Bodies Exploration In Interplanetary Cruise Phase And Rendezvous Phase

With the success of mission “Deep Space1” in America, deep-spaceexploration designed to discovering the mysteries of universe and revealingthe origin of life has attracted research institutes’ attention in many countries.Since deep-space exploration is a long-distance and long-time journey thatrequires accurate and frequent operations to control and change ship’s orbit, ithas high requirements of every key technology in the mission, especiallyautonomous navigation. As a key technology, the autonomous navigation’sstability and precision influences directly the mission’s success and findingefficient methods to conduct autonomous navigation is an important topic.In this paper, the observability analysis methods of the deep spaceautonomous navigation system are studied in the field of geometry andnumerical analysis for interplanetary cruise phase and rendezvous phase. Westudied the feasibility of autonomous navigation method for state estimationbased on camera image information; we built the model of the spaceshipaccording to the kinetic equation and discuss the transition among inertialcoordinate, camera coordinate and image pixel coordinate. Based onasteroids’ orbit parameters, we rebuilt and simulated asteroid orbits. Theautonomous navigation filtering methods during interplanetary cruise phasebased on Sun Line-of-sight vector and asteroid Line-of-sight vector arediscussed in the paper. We also studied about autonomous navigation filteringmethods during rendezvous phase based on Sun Line-of-sight vector andrelative Line-of-sight vector. We used UKF (Unscented Kalman Filter) andEKF (Extend Kalman Filter) for state estimation. We fused the UD factorization into EKF filtering to improve the stability and precision. Wesimulated all these methods by using computer simulation. The study focus onfour aspects:1) The observability analysis methods of the deep space autonomousnavigation system are studied. Considering the environmental characters ofdeep space, the corresponding surfaces of spacecraft position with differentmeasurement information are given utilizing the geometry method. By takingthe partial derivatives of the measurement with respect to each state variable,the observability of the autonomous navigation system is analyzedqualitatively.2) The autonomous navigation scheme of the interplanetary cruise phasebased on Sun Line-of-sight vector is studied. First we discussed the transitionamong inertial coordinate, camera coordinate and image pixel coordinate. Forthe autonomous navigation system utilizing the Sun Line-of-Sight vector, aninitial orbit determination algorithm utilizing three Line-of-Sight vectorsmeasurement is given based on observability degree analysis. After that, theautonomous navigation method introduced the radial velocity is researched.Furthermore, the autonomous navigation method based on the Line-of-Sightvector is studied from the combination of the Sun LOS vector and the opticalinformation, and an autonomous navigation algorithm of information fusionbased on observability degree is given. We studied the EKF based on UDfactorization and UKF to estimate the orbit and then simulated.3) The autonomous navigation scheme of the cruise phase based onasteroid image is studied. Based on asteroids’ orbit parameters, we rebuilt andsimulated asteroid orbits. Navigation measurement model is constructed asthe Line-of-Sight vectors, and after given the selection criteria, the navigationasteroid is chosen according to the observable degree. We used the EKF basedon UD factorization to estimate the orbit and then simulated.4) We studied the EKF based on UD factorization and UKF to estimatethe orbit and then simulated. Considering the ephemeris information of thetarget celestial body, the autonomous navigation method based on the relative Line-of-Sight vector and the Sun Line-of-Sight vector are studied respectively.For the rendezvous phase, the RSEN (Reduced State Encounter Navigation)of the relative motion is given. The relative position vector is deduced underthe different previous situation of the relative velocity, and the relativenavigation method based on the Line-of-Sight vector is researched. We usedthe EKF based on UD factorization and UKF to estimate the orbit and thensimulated.