A BDS/GPS Integrated Positioning Algorithm Based On Kalman Filter and Deep Neural Networks.

Xia Jiang,Qingli Li,Decheng Wei

DOI: https://doi.org/10.1109/CISP-BMEI60920.2023.10373356

2023-01-01

Abstract:As fueled by the advancement of machine learning, the application of neural networks, a branch of machine learning, has been increasingly extensive. Meantime, with the advent of BDS, any position worldwide can be fixed with BDS/GPS. Though BDS/GPS exhibits high accuracy, a more accurate navigation by integrating BDS/GPS positioning is required. Thus, a dual model of integrating BDS/GPS navigation data is introduced in this study. To exploit BDS/GPS and dual model, a novel integrated positioning algorithm combining the Kalman filter and deep neural networks is proposed to simulate the navigation and positioning. The input of our paradigm originates from BDS/GPS and dual model. Stability of raw positioning data can be enhanced by Kalman filtering. Subsequently, a deep neural network is built to integrate the simulations. As revealed from experimental results, stabilization of the input data can enhance reliability of the deep neural network; besides, the proposed data integration algorithm can enhance the accuracy and stability of the navigation and positioning.