Robust motion planning and control of mobile robots for collision avoidance in terrains with moving objects

The present paper studies the problem of control and motion planning in the presence of uncertainty, for a mobile robot subject to state and actuator constraints, that moves along a predetermined path inside a terrain with moving obstacles. The motion of the robot must be as close as possible with respect to time to a nominal velocity profile. The basic idea followed consists in the combination and mutual interaction of a known motion planning method, called minimum interference strategy (MIS), and sliding mode controller. More specifically, the motion planning is performed considering fictitious state and input constraints which are determined taking into account the expected bounds of the declination between the actual and the nominal motions and control laws, in such a way that the actual constraints are not violated despite the uncertainties. The control law parameters must be chosen so that the ranges of motion are restricted as less as possible. The proposed method is computationally efficient and is suitable for on-line implementation. A non-trivial simulation example is included which demonstrates the capabilities of the proposed robust motion controller.