Integration of Planning and Control in Robotic Formations

This paper is devoted to planning and control of a pattern formation of mobile robots when moving between goal points in a known and static environment. Path planning is performed for a reference point in the formation using the modified A* search, coupled with a proposed smoothing technique to generate a feasible trajectory with nonholonomic constraints of mobile robots taken into account. Based on this reference trajectory and the predefined formation configuration in curvilinear coordinates, each robot in the formation computes its trajectory. Formation motion control is then integrated in the proposed framework to derive velocity profiles for robots in the group, taking into account differential geometry of the trajectories. Obstacle avoidance is guaranteed by varying the coordinates of those robots that are likely in collision with obstacles relative to the reference one. Simulation results are presented to illustrate the validity of the proposed framework.