A non-contact method for detecting on-line industrial robot position errors using a microwave Doppler radar motion detector
A non-contact method for detecting on-line industrial robot position errors using a microwave Doppler radar motion detector
复制标题
DOI：
10.1007/s00170-005-2543-y
复制
发表时间：
2006-09
期刊：
影响因子：
--
通讯作者：
Gregory C. Smith;Roger A. Smith
中科院分区：
文献类型：
--
作者：
Gregory C. Smith;Roger A. Smith
工业机器人必须满足其特定应用的定位精度要求。离线校准通常将机器人定位精度提高到大多数工业应用中开环使用所需的水平。对于外部组件要求更高精度的应用通常转向闭环控制或被动顺应性。然而，工业机器人系统通常不监控过程中机器人位置以检测可能导致产品故障、报废、机器损坏和额外成本的机器故障。为了实现更高的操作效率，需要用于监测机器人位置的新的非侵入式、非接触式方法。研究人员开发了一种低成本的方法，用于使用多普勒运动检测器和统计位置误差测量的过程中工业机器人位置监测。该方法在机器人可重复性水平上检测位置误差。
To be useful, industrial robots must meet positioning accuracy requirements for their given applications. Off-line calibration generally improves robot positioning accuracy to levels needed for open-loop use in most industrial applications. Applications that require greater accuracy with respect to external assemblies generally turn to closed-loop control or passive compliance. However, industrial robot systems do not generally monitor in-process robot position to detect machine faults that can lead to product faults, scrap, machine damage, and additional costs. To achieve greater operational efficiencies, new non-invasive, non-contact methods for monitoring robot position are needed. The investigators developed a low-cost method for in-process industrial robot position monitoring using a Doppler motion detector and a statistical position error measure. The method detects position errors at robot repeatability levels.