基于灰色模糊聚类和LS-SVM加工中心的热误差补偿模型

Compensation model for thermal error of machining center based on gray-fuzzy clustering and LS-SVM

为实现数控机床热误差的补偿,提出了基于灰色综合关联度的灰色-模糊聚类算法和最小二乘支持向量机(LS-SVM)对数控机床热误差元素进行优化建模的方法.该方法通过计算各温度测点和热误差数据间的灰色综合关联度,确定灰色相似矩阵,并利用最大树法,得到基于不同水平的聚类结果形成的谱系图,从而确定关键测温点,再利用最小二乘支持向量机方法构建数控机床热误差补偿模型.以MDV-55立式精密加工中心为实验对象进行建模补偿,结果表明,该方法不仅减少了温度传感器的数量,而且机床的加工精度也得到了显著改善.

In order to realize the compensation for the thermal error of numerical control（NC） machine tool,a gray-fuzzy clustering algorithm based on synthetic grey correlation and an optimum modeling method for the thermal error element of NC machine tool with least squares support vector machine（LS-SVM） were proposed.Through calculating the synthetic grey correlation between the temperature measuring points and thermal error data,a gray similar matrix was determined in the proposed method.With the maximum tree method,a hierarchical diagram based on clustering results with different levels was obtained,and thus,the key temperature measuring points were determined.Then the compensation model for the thermal error of NC machine tool was established with a LS-SVM method.The MDV-55 vertical precision machining center was taken as the experimental object to perform the modeling and compensation.The result shows that the proposed method can not only reduce the number of temperature sensors,but also improve the machining accuracy of the NC machine tool obviously.