Application of Machine Learning Methods for Vector of RC States Analysis

Modern local navigation systems have a large number of data sources, which often have low accuracy and a large amount of noise. The Kalman filter can minimize the impact of these noises on the data, but for nonlinear problems this cannot be done due to the difficulty of determining the distribution of the noise. Various machine learning models are used to find such non-trivial dependencies. This paper presents a review of ways to improve the accuracy of the Kalman vector result for robotic systems (RC) local navigation using machine learning methods.