A Learning Servo Motion Controller Based on Cerebellar Model Articulation Controller (CMAC) for Ultra Precision Machine Tools

基于小脑模型关节控制器（CMAC）的超精密机床学习伺服运动控制器

• 批准号: 9322808
• 负责人: Sabri Cetinkunt
• 金额: $ 18万
• 依托单位: University of Illinois at Chicago
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1994
• 资助国家: 美国
• 起止时间: 1994-06-01 至 1997-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9322808&HistoricalAwards=false
• 关键词: Learning; Servo; Motion; Controller; Based

9322808 Cetinkunt The award is for research aimed at improving the accuracy of low speed motion in ultra-precision machine tools by improving control algorithms. The control algorithm is based on the cerebellar model articulation controller architecture and is expected to learn machine characteristics such as friction and backlash, which have a significant effect on machine accuracy at the low speeds. Nanometer level positioning accuracy is sought. Experimental work will be conducted at the National Institute of Standards and Technology and at the grantee institution on a single point diamond turning machine. Ultra-precision machining is needed for components such as metallic and non-metallic mirrors for high power laser focusing, mirrors for compact discs and video-cassette recorders, molds for aspherical lenses, and planarization of silicon wafers. Achievement of the motion accuracies sought here would therefore have a beneficial impact on these manufacturing applications.

该奖项旨在通过改进控制算法来提高超精密机床低速运动精度的研究。该控制算法基于小脑模型关节控制器架构，并期望学习机器特性，如摩擦和间隙，这些特性在低速下对机器精度有重大影响。寻求纳米级的定位精度。实验工作将在国家标准与技术研究所和受资助机构进行，用于单点金刚石车床。高功率激光聚焦的金属和非金属反射镜、光盘和录像机的反射镜、非球面透镜的模具以及硅片的平面化等部件都需要超精密加工。因此，实现这里所寻求的运动精度将对这些制造应用产生有益的影响。