Optimization of Flip-Landing Trajectories for Starship based on a Deep Learned Simulator

TL;DR

This paper introduces a differentiable optimization framework using a neural network surrogate for designing flip-landing trajectories of reusable spacecraft like Starship, enabling efficient gradient-based optimization of complex maneuvers.

Contribution

It presents a novel end-to-end differentiable approach combining deep learning and rigid-body dynamics for spacecraft trajectory optimization, handling nonlinearities and constraints effectively.

Findings

Effective modeling of complex maneuvers with high nonlinearities

Successful integration of neural network surrogate with differentiable dynamics

Potential for future extensions to unsteady aerodynamics and guidance

Abstract

We propose a differentiable optimization framework for flip-and-landing trajectory design of reusable spacecraft, exemplified by the Starship vehicle. A deep neural network surrogate, trained on high-fidelity CFD data, predicts aerodynamic forces and moments, and is tightly coupled with a differentiable rigid-body dynamics solver. This enables end-to-end gradient-based trajectory optimization without linearization or convex relaxation. The framework handles actuator limits and terminal landing constraints, producing physically consistent, optimized control sequences. Both standard automatic differentiation and Neural ODEs are applied to support long-horizon rollouts. Results demonstrate the framework's effectiveness in modeling and optimizing complex maneuvers with high nonlinearities. This work lays the groundwork for future extensions involving unsteady aerodynamics, plume interactions, and intelligent guidance design.