Maneuverability Constraints for Design and Control of Robotic Systems: A Semi-Infinite Programming Approach

This project report presents methods for determining maneuverability constraints for robotic systems. These maneuverability constraints are limits on the acceleration and velocity of the system specified in a global coordinate frame (e.g., a Cartesian reference frame where trajectory planning is typically done) based on torque/force and operating limits in a different reference frame (e.g., the manipulator joint angles where the system dynamics are more easily specified). We approach this problem of determining maneuverability constraints from an optimization perspective. The formulation as a mathematical program results in a generalized semi-infinite programming problem. An algorithm to solve the generalized semi-infinite programming problem is presented for the case when the objective function is a function of a single scalar variable. This algorithm is subsequently applied to two robotic manipulator applications. These examples illustrate the use of the maneuverability results for design and control of robotic systems. Acknowledgments I would like to sincerely thank my advisor Professor Bruce Krogh for the guidance and motivation he provided for me in the course of this work. The other members of the committee, Dr. C.P. Neuman and Dr. Mark Nagurka, also deserve much appreciation for their comments and suggestions. Finally, I would like to thank my family, relatives, and friends for their encouragement, smiles, and laughs that make life so much fun.