Real-Time Calibration of Odometer in Integration with LiDAR and Gyroscope for Map-Aided Positioning during GNSS Outages

Abstract

In recent research, significant efforts have focused on achieving dependable real-time positioning in challenging environments, which is a crucial aspect for the development of various Intelligent Transportation Systems (ITS) applications. Given the limitations of Global Navigation Satellite Systems (GNSS) in suburban and urban areas, where signal blockage is common, there is a growing need for an independent positioning system to provide accurate and continuous location data during GNSS disruptions. Previous studies have explored the combination of Light Detection and Ranging (LiDAR), gyroscopes, and odometer sensors for this purpose. This research builds upon that foundation by introducing a real-time calibration process for odometer readings, leveraging road maps and a road segmentation technique. To evaluate this method, real-world data collected from a moving vehicle was used, incorporating three five-minute simulated GNSS outages. These data were processed in a simulated real-time mode. The results from these tests are promising, showing notable improvements in navigation accuracy. Specifically, the application of the real-time calibration method led to an enhancement in positioning accuracy by 0.9, 1.0, and 0.2 m for each of the GNSS outages, respectively, highlighting the critical role of this calibration process. The performance of the algorithm was improved during the second and third outages with the increased availability of line features. The proposed simpler LiDAR data processing algorithm could achieve mean positional errors of 1.8 and 1.8 m, with maximum errors of 4.0 and 3.8 m, respectively.