Path planning and tracking for robots based on cubic hermite splines in real-time

Abstract

This paper deals with path planning using cubic Hermite splines in real-time control of wheeled robots. The first part describes the filtering and prediction of the robot position. The discrete form of the extended Kalman filter is introduced on the basis of the robot model. The second part shows a computation of trajectories based on Hermite splines. The details of recomputation of the trajectory in each system cycle are mentioned. The velocity profile of the computed trajectory is then determined in order to achieve smooth movement within given physical constraints. A Feedback and feedforward controller is used to control the motion on the calculated trajectory. The paper concludes with a comparison of the trajectory computed by simulation and the trajectory measured on the real robot.