Optimization of GNSS/IMU Positional Accuracy with Kalman Filtering and RTK Corrections for Autonomous Vehicle Applications

Autonomous vehicle technology depends on precise navigation systems that are vulnerable to GNSS signal disruptions. This research investigates the integration of GNSS and inertial measurement unit (IMU) sensor data through a tightly coupled integration (TCI) method using the Kalman filter (KF) and extended Kalman filter (EKF). Combining raw GNSS carrier phase and pseudo-range measurements with IMU data enhances sensor efficiency and resilience to outliers. This study also assesses the effectiveness of real-time kinematic (RTK) corrections with KF and EKF. Our findings demonstrate that the EKF significantly improves positional accuracy by effectively mitigating non-linear errors in the tightly coupled integration of GNSS/IMU measurements. Specifically, EKF reduces the 2DRMS (horizontal) error from 0.780m to 0.719m and the 3DRMS (vertical) error from 0.390m to 0.359m when integrated with RTK. These results highlight the importance of advanced filtering and sensor integration for enhancing navigation accuracy, which is crucial for the safety and performance of autonomous vehicles.