Tool wear prediction method based on the SVM-Clara model
- Autores: Yuzang Tang, Liang Sun
- Localización: Applied Mathematics and Nonlinear Sciences, ISSN-e 2444-8656, Vol. 9, Nº. 1, 2024
- Idioma: inglés
- Enlaces
-
Resumen
To reduce the damage of mechanical parts during machining, a tool wear prediction method based on the SVM-Clara model is proposed. By analyzing the support vector machine (SVM) and Clara algorithm, using regular prediction data or unobservable data, the average dissimilarity of all objects is concentrated, and the characteristics of the overall data are accurately represented. Randomly select data samples from the overall data samples according to a certain proportion, and standardize them to improve the clustering quality. Find the best objective function to minimize the damage function and make the predicted value closer to the actual value. Through experiments, it is proved that the method in this paper can accurately predict the tool wear condition, the mean square error value is 0.03, the prediction method is better, and the production efficiency is ensured.
