Increasing FastSLAM accuracy for radar data by integrating the Doppler information

Abstract

This paper presents a modified FastSLAM approach for the specific application of radar sensors using the Doppler information to increase the localization and map accuracy. The developed approach is based on the FastSLAM 2.0 algorithm. It is shown how the FastSLAM 2.0 approach can be significantly improved by taking the Doppler information into account. Therefore, the modelled, so-called expected Doppler, and the measured Doppler are compared for every detection. Both, simulations and experiments on real world data show the increase in accuracy of the modified FastSLAM approach by incorporating the Doppler measurements of automotive radar sensors. The proposed algorithm is compared to the state-of-the-art FastSLAM 2.0 algorithm and the vehicle odometry, whereas profiles of an Automotive Dynamic Motion Analyzer serve as the reference.