A new correction method for quasi-Keplerian orbits

TL;DR

This paper introduces a correction method using scale factors and iterative algorithms to preserve integral invariants in quasi-Keplerian orbit simulations, improving numerical accuracy in two-body and N-body problems.

Contribution

The paper proposes a novel correction scheme employing Newton's method and SVD to enforce integral constraints, enhancing the accuracy of orbital computations.

Findings

Significantly reduces numerical errors in orbital elements.

Applicable to perturbed two-body and N-body problems.

Improves long-term stability of orbital simulations.

Abstract

A pure two-body problem has seven integrals including the Kepler energy, the Laplace vector, and the angular momentum vector. However, only five of them are independent. When the five independent integrals are preserved, the two other dependent integrals are naturally preserved from a theoretical viewpoint; but they may not be either from a numerical computational viewpoint. Because of this, we use seven scale factors to adjust the integrated positions and velocities so that the adjusted solutions strictly satisfy the seven constraints. Noticing the existence of the two dependent integrals, we adopt the Newton iterative method combined with the singular value decomposition to calculate these factors. This correction scheme can be applied to perturbed two-body and N-body problems in the solar system. In this case, the seven quantities of each planet slowly vary with time. More accurate values can be given to the seven slowly-varying quantities by integrating the integral invariant relations of these quantities and the equations of motion. They should be satisfied with the adjusted solutions. Numerical tests show that the new method can significantly reduce the rapid growth of numerical errors of all orbital elements.