7 Conclusions and directions for further work

 I have presented an efficient method, based on genetic algorithms, for finding the orbital parameters of interacting galaxies, and applied it to simulated galaxies on both hyperbolic and elliptical orbits. Given the centre-of-mass positions of both galaxies, their radial velocities, the scale radii and orientation of their discs, and a matrix of observed (light or mass) density data, in some cases supplemented with velocity data, the GA was able to find the orbital parameters with great accuracy in most cases. In the cases where an orbit could not be found, the failure could be detected rather quickly from the fitness values.

The GA can operate on reduced data sets, in which the inner regions have been blocked out, enhancing the tidal features which are needed for the determination of the orbital parameters.

Even though there have been several simplifications in the test cases used here, the GA method is a promising approach to the type of problems considered in the paper. In order to further improve the method, the possibility of including the scale radii, scale heights, inclinations, and position angles of the two discs in the chromosomes could also be considered. However, as mentioned above, these parameters can often be measured, and the benefits obtained by including them in the chromosomes may not be sufficiently strong to justify the added complexity thus introduced. With further increases in computer speed, another improvement could come from using a simulation code incorporating the self-gravity of the two galaxies. Dark matter haloes could also be added, and if the distributions of dark matter were chosen in such a way as to be easily parametrized by one or a few parameters, it would be possible to add also these parameters to the set of unknowns.

Still, even in its more primitive present state, the GA method is very useful for finding orbital parameters. The output orbital parameters of the best simulation in a GA run can be used as input parameters for an advanced self-gravitating N-body simulation incorporating gas dynamics and dark matter.