Design and Analysis of Haptic Interface and Teleoperator Feedback Systems.

Abstract

This dissertation analyzes feedback design within haptic interface and teleoperator systems to reveal fundamental tradeoffs between design objectives, uncover intrinsic limitations imposed by hardware, and improve existing design practice. The challenge of haptic rendering and teleoperation is to synthesize a realistic mechanical sensation through feedback control while achieving other satisfactory feedback properties including robustness to hardware, noise attenuation, and stability. Special performance requirements and human-in-the-loop stability issues inherent to haptic rendering and teleoperation mean that certain conventional tools for servo-control design are not applicable. This dissertation addresses the gap in applicable theory by applying linear systems analysis to reveal previously unrecognized algebraic and analytic design relationships within haptic rendering and teleoperation. The introduction of distortion as a new performance metric for haptic rendering and teleoperation is a key contribution of this work and leads to a suite of new design relationships and tools.

Important feedback design goals including performance, stability robustness, insensitivity to hardware parameter variations, and noise attenuation present a multi-objective synthesis problem with intrinsic tradeoffs. Furthermore, properties of the hardware including actuator bandwidth limitations, sensor and actuator noise, hardware nonlinearities and lightly damped structural modes constrain the feedback design and achievable goals. The analyses of haptic rendering and teleoperation presented in this dissertation yield relationships that distinguish feasible from infeasible specifications and predict performance as well as other feedback properties that may be expected from a well-tuned controller. Hardware dynamics play a key role in feedback design tradeoffs and limitations. If desired feedback properties are not feasible with given hardware, interpretation of tradeoff relationships and limitations provides direction for hardware re-design.