In this project we plan to use YOLO to code an AI to identify exoplanets. We’ll then use a couple of calculations to find if the planet is potentially habitable. To do this, we first researched different ways of identifying exoplanets manually. We discovered many methods to identify exoplanets such as direct imaging, radial velocity, gravitational microlensing and transit light graphs. We decided to use the transit light graph method as it is relatively simpler than the other methods, requiring less work and/or specific machinery and has been the most successful in discovering new exoplanets.
After choosing our method, we found various websites with telescope data. We ended up using the graphs from Andrew Vanderburg’s website which can be found in the references section. We also had to build on a theory we had on finding if they were habitable.
To do this we first conducted some research on factors that affect the habitability of a planet. Once we knew what makes a planet able to sustain life we were able to use various mathematical formulae to figure out if the planet we found is habitable. The research can be found under their own respective headings and the formulae can be found under the heading “Experimental methods”.
We had to go through lots of problems with the website such as it being down, data issues and buggy code, false measurements and miscalculations but eventually we found a solution to every problem we faced.
|