不可见的
支持向量机
聚类分析
机械加工
刀具磨损
计算机科学
数据挖掘
功能(生物学)
价值(数学)
模式识别(心理学)
人工智能
机器学习
数学
工程类
计量经济学
生物
机械工程
进化生物学
标识
DOI:10.2478/amns.2023.1.00249
摘要
Abstract 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.
科研通智能强力驱动
Strongly Powered by AbleSci AI