An Adaptive Tool-Based Telerobot Control System

Modern telerobotics concepts seek to improve the work efficiency and quality of remote operations. The unstructured nature of typical remote operational environments makes autonomous operation of telerobotic systems difficult to achieve. Thus, human operators must always remain in the control loop for safety reasons. Remote operations involve tooling interactions with task environment. These interactions can be strong enough to promote unstable operation sometimes leading to system failures. Interestingly, manipulator/tooling dynamic interactions have not been studied in detail. This dissertation introduces a human-machine cooperative telerobotic (HMCTR) system architecture that has the ability to incorporate tooling interaction control and other computer assistance functions into the overall control system. A universal tooling interaction force prediction model has been created and implemented using grey system theory. Finally, a grey prediction force/position parallel fuzzy controller has been developed that compensates for the tooling interaction forces. Detailed experiments using a full-scale telerobotics testbed indicate: (i) the feasibility of the developed methodologies, and (ii) dramatic improvements in the stability of manipulator – based on band saw cutting operations. These results are foundational toward the further enhancement and development of telerobot.