Performance improvement of industrial robot trajectory tracking using adaptive-learning scheme

More and more industrial robot operations demand high-accuracy trajectory p erformance which may not be achievable by using conventional PID control. T his paper describes a new adaptive control method with a learning ability i n the repetitive tasks, called the Adaptive-Learning (A-L) scheme. The meth od is based on the proposed theory of two operational modes: the single ope rational mode and the repetitive operational mode. In the single operationa l mode, the control is an adaptive control with a new parameter adaptation law using information from the previous trials. In the repetitive operation al mode, the control is a model-based iterative learning control. The advan tage of the A-L scheme lies in the ability to guarantee convergence in both modes. Theoretical analysis and experimental evaluation on a commercial ro bot demonstrate the effectiveness of the A-L scheme in controlling an indus trial robot manipulator.