Research on data fusion algorithm of GPS and IMU based on Kalman filter, machine learning, and deep learning algorithms

The Global Positioning System (GPS) has the advantages of global coverage and high accuracy, but larger obstacles can affect its positioning accuracy; Inertial Measurement Unit (IMU) is not affected by obstacles, but long-term positioning can cause IMU to accumulate errors, thereby affecting positioning accuracy. On how to improve the accuracy and robustness of GPS and IMU, this article compares the performance of machine learning, deep learning, and Kalman filter estimation algorithms in predicting and optimizing vehicle trajectories, analyzes the mean square error (MSE) and R2 (Coefficient of Determination) of multiple algorithms, and explores the impact of noise on accuracy. Research has shown that the Kalman filter estimation algorithm performs well in dealing with noise in IMU and GPS fusion data, and machine learning and deep learning algorithms also have certain potential in this regard.