Prediction and refinement of ionospheric corrections in radionavigation based on machine learning and Kalman filter using data from electron concentration models and radionavigation measurements

The article discusses a methodology for predicting and refining ionospheric corrections in radio navigation using machine learning and the Kalman filter. The first source of data is the predicted values of the vertical electron density of the ionosphere obtained using machine learning methods, which allows for increasing the accuracy of ionospheric corrections. The second source is the values calculated through the accumulation of radio navigation measurements in real time, which ensures the relevance and efficiency of the data. The proposed methodology allows combining the advantages of modeling and real measurements to improve the accuracy of ionospheric effects correction. The experimental results demonstrate the effectiveness of the approach and its potential for application in modern satellite navigation systems.