Parallel, adaptive, multi-object trajectory integrator for space simulation applications

Abstract

Computer simulation is a very helpful approach for improving results from space born experiments. Initial-value problems (IVPs) can be applied for modeling dynamics of different objects – artificial Earth satellites, charged particles in magnetic and electric fields, charged or non-charged dust particles, space debris. An ordinary differential equations systems (ODESs) integrator based on applying different order embedded Runge–Kutta–Fehlberg methods is developed. These methods enable evaluation of the local error. Instead of step-size control based on local error evaluation, an optimal integration method is selected. Integration while meeting the required local error proceeds with constant-sized steps. This optimal scheme selection reduces the amount of calculation needed for solving the IVPs. In addition, for an implementation on a multi core processor and parallelization based on threads application, we describe how to solve multiple systems of IVPs efficiently in parallel.The proposed integrator allows the application of a different force model for every object in multi-satellite simulation models. Simultaneous application of the integrator toward different kinds of problems in the frames of one combined simulation model is possible too.The basic application of the integrator is solving mechanical IVPs in the context of simulation models and their application in complex multi-satellite space missions and as a design tool for experiments.