Motion planning for cooperating mobile manipulators

We address the problem of motion planning for nonholonomic cooperating mobile robots manipulating and transporting objects while holding them in a stable grasp. We present a general approach for solving optimal control problems based on the calculus of variations. We specialize this approach to solving the motion planning problem and obtaining trajectories and actuator forces/torques for any maneuver in the presence of obstacles. The approach allows geometric constraints such as joint Limits, kinematic constraints such as nonholonomic velocity constraints, and dynamic constraints such as frictional constraints and contact force constraints to be incorporated into the planning scheme. The application of this method is illustrated by computing motion plans for several examples, and these motions plans are implemented on an experimental testbed. (C) 1999 John Wiley & Sons, Inc.