MOTION CONTROL EKF BASED FOR WHEELED AUTONOMOUS VEHICLES

In this paper a kinematical and dynamical closed loop control system for motion control problem of wheeled vehicle using an Extended Kalman Filter (EKF) in the feedback state controller is presented. The asymptotical and the global exponential stability of the tracking errors are proved, where the exponential stability ensures a good robustness with respect to external disturbances violating the nonholonomic constraints of the motion. The EKF filters noises with Gaussian distribution, while the noises which perturb the nonholonomic constraints are not with Gaussian distribution and they have to be compensated by the robust feedback control system. The algorithm robustness and efficiency are validated by simulations carried out with Matlab software, where good performances are obtained both in case of straight-lines and curvilinear trajectory.