SPSOC: Staged Pseudo-Spectral Optimal Control Optimization Model for Robotic Chinese Calligraphy

Abstract

As the executive module of writing tasks, writing control models play a crucial role in robotic Chinese calligraphy. In this study, in contrast to most of the current models that only consider the optimization of writing trajectory, according to the writing characteristics of Chinese calligraphy, a writing control model based on the pseudo-spectral optimal control method is proposed to optimize the writing trajectory and control parameters of the robotic arm. In the proposed model, we describe the robot writing problem as a trajectory optimization problem. The optimal trajectory fitting curves of the skeletons of basic strokes are designed by the least square fitting method. Moreover, the staged pseudo-spectral optimal control optimization model of qibi, xingbi and shoubi is incorporated into the SPSOC. In addition to optimizing the writing trajectory, we also analysis the trajectory control parameters to ensure that the parameters of the control model at each stage are conformed by writing characteristics and rules. Compared with the existing control methods, the results of several experiments prove that the proposed model SPSOC can simulate the writing rules of calligraphy well based on the writing control characteristics.