Learning based Modelling of Throttleable Engine Dynamics for Lunar Landing Mission

TL;DR

This work addresses the computational challenge of accurately modeling throttleable engine dynamics for lunar descent by proposing a data-driven system identification approach. Using high-fidelity propulsion model data, it constructs a sparse MIMO finite-time difference model with physics-informed extended state features and polynomial mappings learned via Lasso, enabling efficient closed-loop guidance and control analysis. The method demonstrates accurate replication of thrust, gas pressure, and propellant masses across diverse excitation inputs, with strong generalization to unseen conditions and quantified uncertainty through Monte Carlo simulations. The resulting surrogate provides a practical tool for rapid-descent simulations and onboard control development, reducing the computational burden of high-fidelity models while maintaining accuracy.

Abstract

Typical lunar landing missions involve multiple phases of braking to achieve soft-landing. The propulsion system configuration for these missions consists of throttleable engines. This configuration involves complex interconnected hydraulic, mechanical, and pneumatic components each exhibiting non-linear dynamic characteristics. Accurate modelling of the propulsion dynamics is essential for analyzing closed-loop guidance and control schemes during descent. This paper presents a learning-based system identification approach for modelling of throttleable engine dynamics using data obtained from high-fidelity propulsion model. The developed model is validated with experimental results and used for closed-loop guidance and control simulations.

Figures (9)

• Figure 1: Propulsion System Schematics nath2017mathematical
• Figure 2: Model Error as function of history length
• Figure 3: Model Error as function of sparsity
• Figure 4: Sinusoidal sweep response
• Figure 5: Step-stair response