Evaluating N-Body Algorithms

OData support
Rajacsics Tamás
Department of Automation and Applied Informatics

Nowadays there are plenty of problems in modern science which entail particles or particle simulations to some degree. The above problems are usually hard, expensive or even impossible to test in a real life situation. For example experimenting with colliding subatomic particles in the Large-Hadron-Collider can be rather expensive thus it is unreachable for most of the scientists. A no-go experiment in real life is to examine galaxy evolution and life cycle on the long run. There are space telescopes to make snapshots of celestial object but those are just momentary pictures which cannot answer all the questions about them. How great it would be if it were possible to simulate those types of actions in a short period of time repeatedly. Fortunately, there is a solution.

Astronomy has a discipline called celestial mechanics. Its main focus is to describe the motions of celestial bodies. With these motion functions systems can be simulated, which involve two, three, more, even millions of celestial bodies/particles. This is the point where, my work, this paper steps in. To do these simulations, algorithms are required to describe those motions. There are plenty of ways to carry these out, and that is what I am striving for in the course of my research.

Finally, I will implement a few of these algorithms to see how they perform in real life simulations and what advantages and disadvantages they have.


Please sign in to download the files of this thesis.