A new strategy of real-time thermal error prediction for a machining center using RBF network

• Main Authors: KUO-AN WU, 吳國安
• Other Authors: Hua-Wei Chi
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/34310623337997140805

碩士 === 大葉大學 === 機械工程研究所碩士班 === 94 === ABSTRACT The major factors that affect the machining accuracy of CNC machine tools include geometric errors and structural thermal deformation during cutting. Although being successfully demonstrated in literatures, thermally-induced error compensation techniques were not widely implemented in industries. Most of the error-predict models use temperatures as their mapping input. These mapping models are very sensitive to missing data and faulty signals. A wrong temperature reading can result in an incorrect prediction of thermal error. This research proposes a new approach for predicting spindle thermal error of a machining center. A new on line measurement system using RENISHAW TS27R tool setting probe and a RBF network model that predicts the thermal error are developed. Instead of using temperature readings, the new model uses cutting conditions as the mapping inputs. As a result, the problems occurred in the traditional temperature-error mapping model can be avoided. The experimental results showed the proposed approach can predict the spindle thermal induced error correctly and cost-effectively.