基于GM-LS-SVM层级模型的数控机床热误差建模

来源期刊:中南大学学报(自然科学版)2016年第12期

论文作者:殷国富 谭峰 殷勤 董冠华 王亮

文章页码:4028 - 4035

关键词:数控机床;热误差建模;灰色模型;最小二乘支持向量机;BP神经网络

Key words:CNC machine tool; thermal error modeling; grey model; least squares support vector machine; BP neural network

摘    要:为了更精确地对数控机床热误差进行预测及补偿以提高其加工精度,针对单独使用灰色模型或最小二乘支持向量机模型进行机床热误差建模的不足,并利用这2种模型在数据不同处理阶段的优点,提出一种基于灰色模型和最小二乘支持向量机层级模型的数控机床热误差建模方法。根据机床关键点温度数据和热误差数据,首先建立多个不同数据序列长度的机床热误差灰色模型作为前处理层,然后把经过前处理层前处理的热误差和实测热误差分别作为最小二乘支持向量机模型的输入和输出,作为后处理层,以进行预测精度校正。利用该方法在一台精密卧式加工中心上进行建模实验,并与单独使用灰色模型、最小二乘支持向量机模型和BP神经网络模型进行预测精度对比分析。研究结果表明:基于灰色模型和最小二乘支持向量机层级模型的数控机床热误差建模方法具有更高的预测精度和更强的泛化能力。

Abstract: In order to predict and compensate thermal error of CNC (computer numerical control) machine tools more accurately, and to make up the shortcomings of single use of grey model (GM) or least squares support vector machine (LS-SVM) model, a hierarchical model (GM-LS-SVM) combining the data processing merits of GM with that of LS-SVM was proposed for thermal error modeling in machine tools. According to the temperature data and thermal error data of the machine tool, first several thermal error GMs with different data sequence length were established as preprocessing layer, and then the preprocessed thermal errors of preprocessing layer and the measured thermal errors were used as input and output of LS-SVM model respectively, to correct prediction accuracy. A modeling experiment was carried out on a precision horizontal machining center, and the prediction accuracies were compared among the proposed GM-LS-SVM hierarchical model, GM, LS-SVM model and BP neural network. The results show that the proposed GM-LS-SVM hierarchical model has higher prediction accuracy and better generalization ability than the other three models in CNC machine tools’ thermal error modeling.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号