基于动态自适应LS-SVM的数控机床热误差建模研究  

Study on Thermal Error Model of NC Machine Tools Based on Adaptive LS-SVM

作　　者：范秋凤[1] 翟雁[2] 邢春芳[1] 石峰[1] 

机构地区：[1]安阳工学院电子信息与电气工程学院,河南安阳455000 [2]安阳工学院机械工程学院,河南安阳455000

出　　处：《机械设计与制造》2017年第6期140-142,146,共4页Machinery Design & Manufacture

基　　金：河南省科技厅科学技术研究重点项目(14A510002);国家自然科学基金(U1204613)

摘　　要：热误差是造成切削加工工件精度低的重要因素,研究机床热误差是提高加工精度的有效措施。为此,综合考虑最小二乘法(LS)、支持向量机(SVM)和动态自适应算法的优势情况下,提出了一种基于动态自适应LS-SVM的数控机床热误差建模方法。为构建热误差模型,以数控机床XK713进行试验,通过温度和位移传感器分别获取机床温度值与主轴变形量,同时通过动态自适应算法,参数能够被优化,以及对所采集的数据进行最小二乘支持向量机建模,从而可得该数控铣床热误差模型。通过与LS热误差建模方法进行对比分析,结果表明:所提出的热误差模型的精度远优于LS模型。该方法为机床热误差建模的研究和应用奠定了基础。The thermal error is an importance influence factor resulting in the low precision of workpieces, so study on the thermal error of machine tools is effective measure to improve the machining precision. Hence, the thermal error model of NC machine tools based on adaptive LS-SVM was proposed through the integration of the LS, SVM and the adaptive algorithn. To establish the thermal error model, taking the milling XK713 for example, the temperature of the machine tool and deflection of the main axis can be acquired by the temperature and displacement sensor. The parameters can be optimized based on adaptive algorithm, and the thermal error model of milling XK713 can be determined through collected data and modeling LS-SVM. Meanwhile, the results indicated that the proposed model had high prediction precision and its possessed obvious advantage compared with LS method. Thus, the proposed method lays a foundation for the study and application of modeling thermal error.

关 键 词：机床 热误差 最小二乘法 动态自适应 支持向量机 

分 类 号：TH16[机械工程—机械制造及自动化] TP205[自动化与计算机技术—检测技术与自动化装置]