A novel FastSLAM algorithm based on Iterated Unscented Kalman Filter

In this paper, we propose a novel FastSLAM algorithm (named as IUFastSLAM) which is based on Rao-Blackwellized Particle Filter (RBPF) framework and uses Iterated Unscented Kalman Filter (IUKF) to estimate the landmark locations. Iterated Unscented Kalman Filter (IUKF) can improve estimation accuracy over the Extend Kalman Filter and Unscented Kalman Filter. The experimental results show that the proposed algorithm has a superior performance in estimation accuracy and prolonged consistency when compared with the FastSLAM2.0 and UFastSLAM algorithms.