基于经验模态分解的直驱进给轴热误差组合预测

Combined Prediction for Thermal Error of Direct-driven Feed Shaft Based on EMD

为对直驱进给轴的热误差进行精确预测,提出一种基于经验模态分解的直驱进给轴热误差组合预测模型。使用经验模态分解,将温升序列重构为波动特征不同的频率项和趋势项。根据重构项的频率特性,分别输入长短期记忆神经网络、支持向量机和自回归滑动平均模型,进行组合训练,以预测直驱进给轴的热误差。试验结果表明,组合预测模型的预测精度达到90.25%以上,最大预测误差控制在1.4μm以内,预测效果优于普通单项预测模型。

In order to accurately predict the thermal error of the direct-driven feed shaft,a combined prediction model of the thermal error of the direct-driven feed shaft based on empirical mode decomposition was proposed.Using empirical mode decomposition,the temperature rise sequence was reconstructed into frequency term and trend term with different fluctuation characteristics.Based on the frequency characteristic of reconstructed term,reconstructed term was input the long short-term memory neural network,support vector machine and autoregressive moving average model respectively for combined training,to predict the thermal error of the direct-driven feed shaft.The experimental results demonstrate that the prediction accuracy of the combined prediction model is over 90.25%,and the maximum prediction error is less than 1.4μm.The prediction effect of this model is much better than that of ordinary single-term prediction model.