Dynamic Programming Techniques for Planar Orbital Transfer of Low Earth Orbit Satellites

TL;DR

This paper applies dynamic programming to optimize planar orbital transfers of Low Earth Orbit satellites, validating numerical schemes and exploring computational speed-ups as a step towards full 3D low thrust engine applications.

Contribution

It introduces a dynamic programming approach for orbital transfer optimization in a simplified planar model, aiming to extend to full 3D low thrust scenarios.

Findings

Validated numerical scheme in planar case

Assessed accuracy of the dynamic programming method

Explored techniques to accelerate computations

Abstract

In this paper, we present an application of the optimal control theory to orbital transfer of Low Earth Orbit satellites. The optimal control problem is treated with Dynamic Programming techniques which require solving the Hamilton--Jacobi--Bellman equation on a suitable state space, with the reconstruction of the optimal controls made in the form of a static feedback. In order to validate the numerical scheme without the complexity of the full model, this first study sets the problem in planar form, thus working in a four-dimensional state space. We will study various techniques to speed up the computation, and assess the accuracy of the numerical solution. This project is born from the attempt of evaluating and applying direct method of optimal control techniques based on Dynamic Programming as a complementary approach to the well known indirect methods, as Pontryagin or Lawden. In particular, the final aim is to treat the case of low thrust engines from real use cases, in the full 3-D problem.