Homography-based trajectory tracking of industrial manipulator with iterative learning control

In this paper, a repetitive assembling task is considered for industrial manipulator, where vision-based precise trajectory tracking is necessary. Conventional visual servoing methods suffer from their poor dynamic performance, while the proposed open-plus-closed loop framework solved this problem by iterative learning. System error is defined by the directly measurable homography matrix and feature points, without any homography decomposition. Besides, in the proposed controller, image Jacobean is not used so that no singularity problem exists. System convergence is analyzed and simulation results are given to demonstrate the effectiveness, i.e. fairly accurate trajectory tracking is achieved.