Research On Trajectory Planning And Guidance Algorithms For Pinpoint Mars Landing

Pin-point landing is the inevitable trend and is demanded for the future Mars exploration mission. The level of guidance and control during atmospheric flight determines the future mission’s landing accuracy, even the success of the mission. However, traditional guidance and control method during atmospheric flight could not satisfy the need of the future Mars exploration missions, which demand pin-point landing on the Mars. Therefore, it is necessary to develop a new generation of guidance and control algorithms during atmospheric flight for the future Mars exploration activites.With the supports of the 973 Program “Research on GNC for Precise Landing on Planets” and the National Natural Science Foundation of China “Research on Autonomous Navigation Theory and Method for Deep Space Exploration”, this dissertation deeply studies the trajectory optimization and guidance method for Mars pin-point landing, which aims to the error sources that affect Mars landing accuracy. The main contents of this dissertation are as follows:First, researches on design of the lander’s sates at entry point for Mars landing are carried out. In order to reduce the effect induced by the navigation error uncertainty on the lander’s states at parachute deployment point, the sensitivity function is introduced to propose a robust lander’s states design method considering the navigation uncertainty. However, there needs to map the uncertainties at entry point to the states uncertainties at deployment point through the entry dynamic system, and the computation of the relationship is time comsuming. Therefore, a spectral decomposition based method of landing site uncertainty analysis for Mars entry is proposed. Futhermore, to eliminate the error of this method caused by long time integration, a spectral base revision method is given, and an adaptive structural spectral decomposition based method is proposed. The computational accuracy and efficiency is improved.Second, longitudinal guidance method for Mars entry is explored. The traditional acceleration profile based guidance method is improved. The improvement includes two aspects. First, the improved boundary acceleration-energy profile generation method is proposed, which based on investigating the character of drag acceleration trajectories for landing on higher elevation, and the proposed approach could meet the requirement of landing on higher elevation aera. Second, a sliding mode observer is designed to estimate and compensate the trajectory tracking system parameters with uncertainty, and the robustness and tracking accuracy of the acceleration-energy profile tracking controller is improved.Subsequently, three-dimensional guidance algorithm during atmospheric entry for Mars landing is investigated. In the light of the advantages of pseudo-spectral method in the trajectory optimization, the Radau pseudo-spectral method is used to generate the three dimensional trajectory. Considering the nonlinear intensity and the fitting precision of the radau polynomial to the entry trajectory, a piecewise interval and nodes number adusting strategy is given. Then, the analytical jacobian matrix of the cost and constraints function is derived. Therefore, three-dimensional trajectory generation method is proposed. Furthermore, the relationship between system states and inputs is investigated using Lie derivative, and a trajectory tracking method with control constraints is put forward based on predictive control theory.Finally, powered descent guidance methods for Mars landing are studied. Reduced dynamic model which can be used in guidance algorithm design is obtained by investigating the characters of Mars powered descent dynamics. According to the actual project needs, two optimal models are proposed. Based on the reduced dynamic model and the optimal models, the relationship between co-states, thrust magnitude, and thrust vector is discussed. And then two methods are given to solve the optimal models. Additionally, the analytical solutions for state and co-state differential equations are derived by taking advantage of the relationship between co-states, thrust magnitude, and thrust vector, which could improve the computational efficiency largely. Finally, the powered descent guidance algorithm flow is concluded.