基于遗传算法的运动学误差标定
Kinematics Error Calibration Based on Genetic Algorithm

为了提高空间对接试验平台的对接精度,对其进行了误差分析及标定研究。在并联机构运动学正解和逆解的基础上,利用全微分法推导出了并联六自由度机构的运动学误差标定模型。为了更加有效地对各个误差源进行误差辨识,提出了一种基于遗传算法以矩阵条件数为优化目标、对并联机构的标定测量点进行优化筛选的方法,以降低测量噪声对标定结果的影响。当采用较少的测量数据时,仍然能够取得较好的标定效果,具有较好的抗干扰能力。最后,利用数值仿真算例验证了该方法的有效性。
In order to improve the accuracy of space docking platform, it's necessary to analyze the errors and research the error calibration. Based on forward kinematics solution and inverse solution model, the kinematics error calibration model of 6-DOF parallel mechanism is established with total differential method. For the sake of identifying the value of each error more effective, a new error calibration method based on the genetic algorithm to search the optimal measuring points for calibration is proposed to reduce the effect of measurement noise during the calibration, in which the matrix condition number as the optimization objective. The results show that the calibration is effective even with less measurement points; it has the better anti-noise capability. At last, the effectiveness of proposed method is validated with numerical simulation method