Adaptive Guidance with Real-Time Performance Compensation for Mars Ascent Vehicles under Local Degradation
- Paper ID
97058
- author
- company
School of Astronautics, Beihang University; Lunar Exploration and Space Engineering Center; Deep Space Exploration Laboratory (Tiandu Laboratory)
- country
China
- year
2025
- abstract
The Mars Ascent Vehicle (MAV) mission on Mars faces challenges such as imprecise atmospheric models and strong local climate uncertainties, which reduce the reliability of prior optimal trajectory predictions. At the same time, the MAV's fuel consumption significantly impacts Earth launch missions, raising high demands for fuel optimization. Due to the high computational cost of solving optimal trajectories with aerodynamic drag, MAV missions tend to rely on trajectory tracking guidance or other simple closed-loop guidance strategies, all of which suffer from low efficiency and poor fuel utilization. In addition, the robustness of the guidance strategy is crucial. To address this, this paper proposes an indirect-method-based guidance algorithm using Fast Local Iteration (FLI) to solve the optimal control problem with aerodynamic drag at near-analytical speed. Firstly, a scaling factor {\it k} is introduced into the guidance law, recalculated at each cycle, simplifying the guidance law into a zeroth-order problem that can be solved analytically within each cycle. This process is referred to as local degradation. The introduction of {\it k} is the core innovation of this paper. {\it k} can be understood as a continuation of the zeroth-order problem, corresponding to the {\it k}-continuation of the guidance command solution at the current moment. The optimal solution of guidance is then transformed into the selection of the optimal {\it k}. For scenarios involving atmosphere, the initial value of the aerodynamic drag term is introduced using an estimation algorithm based on the nominal trajectory, with performance compensation updated in real-time through Parameter Adaptive Control (PAC). Secondly, the FLI algorithm is developed to iteratively solve for an appropriate {\it k} to find the best matching solution for each cycle, with detailed theoretical reasoning. Finally, the new algorithm is validated through Monte Carlo simulations with various error scenarios and compared with trajectory tracking guidance. The simulations demonstrate that this guidance strategy inherits the excellent properties of iterative guidance, with strong robustness, which is crucial for mission success on the unknown and distant Martian surface. Additionally, its rapid computation speed and near-optimal fuel consumption are significant advantages.