Comparing the Performance of Different Methods for Estimation in Inertial Navigation Systems

Abstract

There are many positioning systems available today. The most prominent of these sys-tems is the Inertial Navigation System, which is increasingly preferred because it works with its own internal system independent of external stimuli. This system detects posi-tion, orientation and velocity information by means of accelerometer and gyroscope sensors. Using this information, it is possible to make predictions for the next position, orientation and speed with various algorithms.
 In the studies conducted so far, Kalman Filter algorithms have been predomi-nantly used for prediction. In this study, Long Term-Short Memory (LSTM) neural network architecture, Bidirectional Long Short-Term Memory (BLSTM), Gated Recur-rent Unit and Kalman Filtering methods, which are among the deep learning algorithms that have proven themselves as prediction algorithms, are examined in detail and a comparative study is presented.
 Considering the findings obtained in the simulation studies detailed in the thesis, it is revealed that Gated Recurrent Unit GRU estimation algorithm shows the best perfor-mance with an RMSE value of 2.5414. LSTM, BLSTM methods achieved an RMSE val-ue of 2.5547, 2.7592 respectively. In the studies conducted through Kalman filter, one of the traditional estimation methods, an RMSE value of 2.9322 was obtained.