Fine motion control of robot manipulators in deburring applications utilizing cutting tool signals

The feasibility of exclusively utilizing cutting tool signals as an alternative feedback technique to enhance robotic machining of complex precision components is evaluated. A hybrid fine motion control scheme based on sophisticated models of the cutting tool dynamics and the nonlinear cutting process is proposed for accommodating the robot trajectory to positional inaccuracies in automated deburring and chamfering applications. Artificial neural networks as general approximators in conjunction with correlation analysis methods are employed to establish the nonlinear mapping of the measurable cutting tool quantities to the depth of engagement during the cutting operation. A superordinate adaptation strategy is developed to adjust the direction of motion correction and the nominal robot path to the local workpiece displacement. The effectiveness and performance of the proposed fine motion control scheme are demonstrated by some robotic deburring and chamfering experiments of fibre-reinforced plastics with a high-speed cutting tool