基于HPO-ELM的电主轴热误差建模
Thermal Error Modeling of Motorized Spindle Based on HPO-ELM
张珂 1徐鹏 2王展 2王子男2
作者信息
- 1. 沈阳建筑大学机械工程学院,辽宁沈阳 110168;沈阳工业大学机械工程学院,辽宁沈阳 110870
- 2. 沈阳建筑大学机械工程学院,辽宁沈阳 110168
- 折叠
摘要
电主轴转子热误差是导致机床加工误差最主要的因素.为了准确预测电主轴热误差,开展了基于HPO-ELM的电主轴热误差建模研究,采集电主轴温升和热误差数据,利用Kmeans++聚类结合灰色关联度理论对其进行分析,提出基于猎人猎物算法优化极限学习机的热误差建模方法,最后建立电主轴热误差预测模型.对电主轴热误差进行仿真分析,并将所提方法与ELM、GA-ELM方法进行热误差预测准确性对比.结果表明:相比ELM、GA-ELM方法,基于HPO-ELM建立的热误差模型具有更高的准确性,预测精度可达98.08％,预测残差值小于1 µm,证明了此模型具有更好的预测精度和鲁棒性.
Abstract
Thermal error of motorized spindle rotor is the main factor leading to machining error of machine tool.In order to accu-rately predict the thermal error of motorized spindle,the thermal error modeling research based on HPO-ELM was carried out.The tem-perature rise and thermal error data of motorized spindle were collected,and they were analyzed using Kmeans++clustering and grey re-lational degree theory.The thermal error modeling method of extreme learning machine based on hunter-prey algorithm was proposed.Fi-nally,the thermal error prediction model of motorized spindle was established.The thermal error of motorized spindle was simulated and analyzed,and the accuracy of thermal error prediction was compared with that of ELM and GA-ELM methods.The results show that compared with ELM and GA-ELM methods,the thermal error model based on HPO-ELM has higher accuracy,the prediction accuracy can reach 98.08％,and the prediction residual value is less than 1 pm,which proves that the model has better prediction accuracy and robustness.
关键词
电主轴/热误差建模/猎人猎物算法/Kmeans++聚类/灰色关联度分析Key words
motorized spindle/thermal error modeling/hunter-prey algorithm/Kmeans++clustering/grey relational analysis引用本文复制引用
出版年
2024
NETL
NSTL
万方数据