Implementation issues and experimental studies of adaptive robust controllers for robot manipulators

This paper summarizes recent advancements on the adaptive robust control (ARC) of robot manipulators. Some of the critical implementation issues are discussed. These issues include: 1) the selection of suitable ARC controller structures for reducing online computation time and the effect of noisy velocity measurement; 2) online desired trajectory generation and initialization for a better transient performance; 3) the tuning of controller parameters in implementation for achieving high performance; and 4) the selection of the bounds of physical parameters and controller parameters to alleviate the effect of control saturation. Comparative experimental studies done on a SCARA type direct-drive robot are presented to illustrate the advantages and the drawbacks of qualitatively different adaptive and robust control algorithms