Path following of an omni-directional four-wheeled mobile robot

Abstract

Navigation of omni-directional four wheeled mobile robot usually is done by trajectory tracking controller. But, in this article two methods of path following have been presented to reach this purpose. Path following is another navigation approach that is solved some of trajectory tracking navigation systems difficulties like ‘turning back’. Path following navigators usually are designed by two different viewpoints, geometrical approaches or control approaches which both are utilized in this paper. The main contribution of this paper is to present tow simple methods with acceptable navigation performance. Based on the simulation results, it is proofed that, the presented methods are able to navigate the robot on desired paths even in hard situations like measurement delay and external disturbance. Also, advantages of each method is presented based on comparison of simulation results. Another contribution of this work is using dynamic model of the robot instead of kinematic model which is usually is used in path following articles. To reach this goal, at the first step, a Linear Quadratic Regulator (LQR) controller is designed as low level controller, then path following navigator is mounted on high level controller. Note that, weighting matrix of LQR controller (matrix Q) is calculated by Genetic Algorithm to achieve the best possible time performances.