Stochastic Runge-Kutta-Munthe-Kaas Methods in the Modelling of Perturbed Rigid Bodies

In this paper we present how nonlinear stochastic Ito circumflex accent differential equations arising in the modelling of perturbed rigid bodies can be solved numerically in such a way that the solution evolves on the correct manifold. To this end, we formulate an approach based on Runge-Kutta-Munthe-Kaas (RKMK) schemes for ordinary differential equations on manifolds. Moreover, we provide a proof of the mean-square convergence of this stochastic version of the RKMK schemes applied to the rigid body problem and illustrate the effectiveness of our proposed schemes by demonstrating the structure preservation of the stochastic RKMK schemes in contrast to the stochastic Runge-Kutta methods.