Thermal Error Modeling of Machine Tools Using Neural Networks
Alice V. Ling
This paper describes an approach for the enhancement of machine tool accuracy. A prediction of the thermally-induced displacement error of the cutting tool with respect to the workpiece is used for corrective action. Recently, researchers have increased machine tool accuracy by applying error compensation using modeling techniques such as multiple regression analysis or neural networks. Thermal displacement errors are not static in nature, i.e., the machine tool stores thermal effects (thermal memory) during the course of operation. To study the possibility of improving models by attempting to capture thermal memory effects in a model, various types of dynamic and static models were applied to existing displacement data from a turning machine. Comparisons are made between the following: (1) neural network and regression models, and (2) inclusion and exclusion of time delay terms (dynamic vs. static). Results are shown and conclusions made on generalization of the models.
Proceedings for an International Multidisciplinary Conference Intelligent Systems: A Semiotic Perspective
Thermal Error Modeling of Machine Tools Using Neural Networks, Proceedings for an International Multidisciplinary Conference Intelligent Systems: A Semiotic Perspective, Gaithersburg, MD
(Accessed February 21, 2024)