Adaptive Kalman Filter Based LiDAR aided GNSS/IMU Integrated Navigation System for High-Speed Autonomous Vehicles in Challenging Environments

초록

This paper presents an advanced localization method for autonomous vehicles using an adaptive Kalman filter integrated with LiDAR-aided GNSS/IMU systems, addressing the challenges of GNSS reliability and precision in urban high speed driving scenarios. This approach combines multiple sensors, including GNSS, IMU, and LiDAR, with adaptive estimation method to enhance the accuracy of position estimation. The adaptive Kalman filter effectively counters the limitations of GNSS by employing fault detection and exclusion and utilizing LiDAR-based localization method for real-time correction in case of GNSS errors. Experiments conducted with generated critical GNSS errors in a test vehicle driving in urban environments, demonstrate the improved reliability and accuracy in false alarm detection and position estimation of the system. This research contributes to autonomous driving technology, offering a robust solution for precise localization in dynamic urban environments, and is adaptable to a wide range of unmanned mobility systems, signifying a considerable advancement in the field of autonomous navigation. © 2024, Institute of Navigation