Triple GPS Localization

Abstract

The purpose of this thesis is to increase the positional accuracy of Global Position System (GPS) modules using an artificial intelligence algorithm. Three basic and identical GPS modules were setup in an equilateral triangle formation with side lengths of one meter. The triangle was placed in four separate locations where approximately 700 GPS data points were collected per GPS module. The data was then analyzed and tested using a Gaussian distribution. In the best-case scenario, the Triple GPS Localization algorithm was able to improve the localization accuracy by 82%. This shows that utilizing a machine-learning algorithm can improve the positional accuracy of any GPS module.