Waypoint Constrained Multi-Phase Optimal Guidance of Spacecraft for Soft Lunar Landing

Abstract

A waypoint constrained multi-phase nonlinear optimal guidance scheme is presented in this paper for the soft landing of a spacecraft on the Lunar surface by using the recently developed computationally efficient Generalized Model Predictive Static Programming (G-MPSP). The proposed guidance ensures that the spacecraft passes through two waypoints, which is a strong requirement to facilitate proper landing site detection by the on-board camera for mission safety. Constraints that are required at the waypoints as well as at the terminal point include position, velocity, and attitude of the spacecraft. In addition to successfully meeting these hard constraints, the G-MPSP guidance also minimizes the fuel consumption, which is a very good advantage. An optimal final time selection procedure is also presented in this paper to facilitate minimization of fuel requirement to the best extent possible. Extensive simulation studies have been carried out with various perturbations to illustrate the effectiveness of the algorithm. Finally, processor-in-loop simulation has been carried out, which demonstrates the feasibility of on-board implementation of the proposed guidance.

Robust Bang-Off-Bang Low-Thrust Guidance Using Model Predictive Static Programming

“…6) MPSP-CT is superior to TPBVP-CT if NI ≤ 25. However, the real-time capability should be further verified using processor-in-loop simulations [37]. Overall, evidence is given that the developed bang-off-bang MPSP method is effective in providing robust guidance laws.…”

“…Future works will focus on 1) including the fuel consumption into MPSP design to further improve its performance and robustness, 2) reducing the computational time and iterations especially when the thrust sequence changes, and 3) validating the real-time capability through processor-in-loop simulations [37].…”

Review of advanced guidance and control algorithms for space/aerospace vehicles

“…Linear covariance-based MPC (LCMPC) [106] Model predictive static programming (MPSP) [102] Quasi-spectral model predictive static programming (QSMPSP) [104] Model predictive convex programming (MPCP) [107] Sampling-based stochastic model predictive control (SSMPC) [101] Tube-based robust model predictive control (TRMPC) [100] Linear pseudospectral model predictive control (LPMPC) [108] poor convergence due to the existence of higher-order linearization errors and artificial infeasibility [113,114].…”

“…The second path focuses on alleviating the computational burden of the online optimisation process, thus making the algorithm more suitable for practical guidance and control tasks. This leads to the development and application of model predictive static programming (MPSP) [102] and its various enhanced versions [103][104][105].…”

Linear Quadratic Tracking With Reinforcement Learning Based Reference Trajectory Optimization for the Lunar Hopper in Simulated Environment

“…One key difference between lunar hoppers and landers exists in the approaching phase: lunar landers essentially do not need a maneuver to avoid obstacles if they successfully identify them and select an appropriate trajectory to follow while they approach from a high altitude [14]. That said, waypoint-based trajectory design based on the aforementioned Apollo or ZEM/ZEV laws, such as [15], works well. On the other hand, lunar hoppers need more continuous and dynamic trajectory (re)planning to avoid different-sized obstacles as they fly over at a lower altitude and, as a result, alternative guidance methods are required.…”