Iterative-learning integral-plus-proportional control of a flexible swing arm system for position trajectory tracking

Abstract

In the LED packaging industry, the position trajectory tracking of a flexible swing arm system is influenced by the unknown disturbances and parametric uncertainties. In this paper, an iterative-learning integral-plus-proportional (IP) controller is proposed to enhance the efficiency and accuracy of the flexible swing arm system. First, online prediction mechanism is incorporated with the iterative-learning automatic tuning to obtain an optimum of the controller parameters, referred to as predictive iterative-learning control. Then, the adaptive iterative-learning IP controller is derived using the proposed predictive iterative-learning control method. To achieve faster convergence, virtual reference feedback tuning is used to obtain the initial parameters of IP controller. In this method, only input/output measured data of the controlled plant are fully utilized by means of dynamic linearization technique. Moreover, the proposed tuning method can optimize the controller online using experimental data during normal system operation. The convergence and stability properties of the closed-loop system are analyzed. Finally, simulations and experiments on real-time flexible swing arm system show the effectiveness of the predictive iterative-learning IP control method.