In Navigation from LSE to AI. Application of RNN to GA Based BGPS and RTK-B Positioning

We have developed instant positioning method BGPS™ and RTK-B using an application of genetic algorithms (GA), first mentioned in [1], and then detailed in [2],[3]. The integration of genetic algorithms for carrier ambiguity resolution and compact continuum transformation algorithms for computational complexity reduction allows for the attainment of more accurate and reliable positioning solutions with reduced initialization times and elevated integrity (see [4] and [5]) In this paper we research new level of AI (artificial intelligence) application for navigation by introducing an enhancement of the poisoning algorithms with recurrent neural networks (RNN) to improve GA convergence and paving a way for AI application for navigation in the future.