Application of the machine learning regression algorithms to optimization of the CMM measurement strategies
Artykuł w czasopiśmie
MNiSW
70
Lista 2024
| Status: | |
| Autorzy: | Rzeczkowski Jakub, Samborski Sylwester, Czajka Aleksander, Kłonica Mariusz, Janik Mateusz, Krzepek Konrad, Sobecki Piotr, Besztak Marek |
| Dyscypliny: | |
| Aby zobaczyć szczegóły należy się zalogować. | |
| Rok wydania: | 2025 |
| Wersja dokumentu: | Drukowana | Elektroniczna |
| Język: | angielski |
| Numer czasopisma: | 11 |
| Wolumen/Tom: | 36 |
| Numer artykułu: | 115002 |
| Impact Factor: | 3,4 |
| Web of Science® Times Cited: | 0 |
| Bazy: | Web of Science |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | NIE |
| Abstrakty: | angielski |
| In this paper, the optimization of coordinate measuring machine (CMM) measurement strategies by using machine learning regression algorithms was investigated. The study analyzed the impact of key strategy parameters (mainly the measurement speed). On measurement accuracy and repeatability by using the Zeiss Contura equipped with an active probe head. Experimental data were used to develop regression models for optimizing the CMM strategy towards limitation of measurement deviations. Currently, parameter selection relies on manufacturer recommendations and operator input. The novelty of this research lies in applying machine learning techniques to automate this process. Three different regression algorithms; the linear regression, the support vector regression, and the decision tree regression were evaluated by using a dataset consisted of sixty measurement points. In addition, the polynomial feature transformations was applied to input data in order to capture nonlinear relationships. Models were assessed by using the 10-fold cross-validation and performance metrics such as the mean absolute error and the root mean squared error. Results indicated that polynomial transformations improved model performance. The linear regression achieved the lowest errors and best balance between accuracy and models’ complexity. The support vector regression showed strong generalization, while the decision tree regression exhibited overfitting. Learning curves and residual plots confirmed that regularization improved decision tree performance. Obtained outcomes demonstrate that machine learning algorithms can effectively optimize the CMM measurement strategies, reducing operator dependency and enhancing efficiency. Future work should expand the dataset and explore additional algorithms to refine prediction accuracy and parameter selection. |
