关键词: 并联机构, 自标定, 误差建模, 参数辨识, 计算机仿真

Abstract: As it is difficult to realize full closed-loop feedback control for parallel kinematic machines (PKM), their kinematic calibration becomes one effective means of significant economic value to enhance PKM precision. In general, kinematic calibration consists of four steps, i.e. error modeling, measurement, identification, and compensation. By taking a 5degree of freedom (DOF) reconfigurable hybrid robot named Tricept as an object of study, this article presents a kinematic calibration method based on joint sensors. The first order error function is formulated, which maps the measured data and the geometric source errors affecting the compensable pose errors. Based on the error mapping function, the identifiability of geometric parameters of Tricept is investigated. The validity and effectiveness of the proposed approach is verified by computer simulation.

Key words: parallel kinematic machines, self-calibration, error model, parameter identification, computer simulation