A genetic algorithm approach to characterising post-AGB circumstellar environments

The topic of this thesis is the characterisation of the circumstellar environments about AGB and post-AGB stars via high resolution optical (mid-infrared) interferometry. Interferometric observations provide the required high angular resolution to study of these objects, and the parametric constraints of such regions provide insight into the evolutionary pathways that have given rise to these astrophysical objects. This will helps us to reach the goal of understanding the stellar pathways that turn AGB stars into white dwarfs. In this thesis we focus on the development and application of a systematic method of fitting optical interferometric data products, in a quantitative, fully reproducible and non-subjective fashion. Such methods improve upon previous attempts to solve the inverse optimisation problem that exists between the interferometric data outputs and the radiative transfer models that are needed to give quantitative environmental descriptions. By running a number of search heuristics (genetic algorithms) we generate probability density functions associated with each parameter, and reveal parametric correlations, degeneracies, confidence intervals, as well as final best fit solutions.By applying the algorithm to 4 nebulae for which interferometric data had been obtained previously, we confirm the existence of dusty tori. The parameters we derive point to the existence of post-AGB stars at their core. These data, together with other observations from the literature lead us to conclude that the collimated nebulae observed must be the product of a strong binary interaction, where jets have shaped the nebulae or where the collimated nebulae derive from a common envelope ejection of the entire envelope