A random-profile approach for trajectory planning of wheeled mobile robots

We present a random-profile approach to treat the optimal free-trajectory planning problem for nonholonomic wheeled mobile robots subjected to move in a constrained workspace. This versatile method is based on a simultaneous search for the robot path and for the time evolution on this path. It handles obstacle avoidance issues while considering kinodynamic constraints (bounded velocities, accelerations and torques). It may be applied to treat problems with various forms of optimization criteria involving travel time, efforts and power. Numerical results, obtained via the simulated-annealing technique of optimization, are presented for two- and three-wheel mobile robots and are compared to those available in the literature.