Autonomous Vehicle State Estimation using a LPV Kalman Filter and SLAM

This paper presents an optimal approach for state estimation and Simultaneous Localization and Mapping (SLAM) correction using Kalman gain obtained via Linear Matrix Inequality (LMI). The technique utilizes a Linear Parameter Varying (LPV) represention of the system, which allows to model the complex non-linear dynamics in a way that linearization is not required for the estimator or controller design. In addition, the LPV polytopic representation is exploited to obtain a real-time Kalman gain, avoiding expensive optimization of LMIs at every step. The estimation schema is integrated with a Non-linear Model Predictive Control (NMPC) in charge of controlling the vehicle. For the demonstration, the approach is tested in the simulation and for the practical validity, a small-scale autonomous car is used.