plug-and-play positioning error compensation model for ripple suppressing in industrial robot polishing

Industrial robot-based polisher has wide applications in the field of optical manufacturing due to the advantage of low cost, high degrees of freedom and high dynamic performance. However, the large positioning error of the industrial robot can lead to surface ripple and seriously restrict the system performance; however, this error can only be inefficiently compensated by measurement before each processing at present. To address this problem, we discovered the period-phase evolution law of positioning error and established a double sine function compensation model. In the self-developed robotic polishing platform, the results show that the Z-axis error in the whole workspace after compensation can be reduced to ±0.06mm, which reaches the robot repetitive positioning error level; the Spearman correlation coefficients between the measurement and modeling errors are all above 0.88. In the practical polishing experiments, for both figuring and uniform polishing, the ripple error introduced by positioning error is significantly suppressed by proposed model under different conditions. Besides, the Power Spectral Density (PSD) analysis has shown a significant suppression in the corresponding frequency error. This model gives an efficient plug-and-play compensation mode for robotic polisher, which provides new possibilities for further improving robotic processing accuracy and efficiency.

基于工业机器人的抛光设备在光学制造领域应用广泛，其具备成本低廉、自由度高与动态性能优异的优势。然而，工业机器人存在较大的定位误差，会引发表面波纹问题，严重制约系统性能；而当前仅能通过每次加工前开展测量的方式低效补偿该误差。为解决这一问题，我们探明了定位误差的周期-相位演化规律，并建立了双正弦函数补偿模型。在自研的机器人抛光平台上，实验结果显示：补偿后全工作空间内的Z轴误差可降至±0.06mm，达到了机器人重复定位误差的水平；测量误差与建模误差间的斯皮尔曼（Spearman）相关系数均高于0.88。在实际抛光实验中，针对修形与均匀抛光两种工况，所提模型可在不同条件下显著抑制由定位误差引发的波纹误差。此外，功率谱密度（Power Spectral Density，PSD）分析表明，对应频率段的误差得到了有效抑制。该模型为机器人抛光设备提供了一种高效的即插即用补偿方案，为进一步提升机器人加工的精度与效率提供了新的可能性。