Dual Extended Kalman Filter Based State and Parameter Estimator for Model-Based Control in Autonomous Vehicles

To enable effective model-based control for autonomous vehicles, accurate vehicle states and model parameters are required. Most of the state estimation and model-based control consider a linear time-invariant model with fixed system parameters by using their approximate values, such as the weight of the vehicle and tire cornering stiffness. However, these parameters may vary significantly due to the highly nonlinear behavior of the tires of the vehicle and changes in the driving environment. In this work, we propose a state and parameter estimation approach based on the dual extended Kalman filter (DEKF) to obtain accurate states and time-varying parameters. The system is modeled as a linear parameter varying system. New formulations of parameter correction for general DEKF are proposed to achieve faster convergence. Results of simulations and experiments on a real vehicle are included to demonstrate the effectiveness of the proposed method and its benefits for model-based control approaches.