基于SO-ELM的数控机床进给系统热误差分析

Thermal Error Analysis of CNC Machine Tool Feed System Based on SO-ELM

为了对数控机床进给系统热误差进行更加精确的预测,提出一种基于蛇优化(SO)算法和极限学习机(ELM)的数控机床进给系统热误差预测模型SO-ELM。利用模糊c均值聚类(FCM)和灰色关联度分析(GRA)筛选出进给系统的关键测温点;通过蛇优化算法优化极限学习机的输入层权重和隐藏层偏置,利用关键温度测点的温升数据和热误差数据构建SO-ELM热误差预测模型。为验证模型的准确性和适用性,与基于SSA-BP和LSMT的热误差预测模型进行对比分析,结果表明SO-ELM模型预测结果的均方根误差和决定系数均优于SSA-BP和LSTM模型,能够更精准地对机床进给系统热误差进行预测,为机床热误差预测补偿提供一种新的思路。

In order to predict the thermal error of CNC machine tool feed system more accurately,a numerical control machine tool feed system thermal error prediction model SO-ELM based on snake optimization(SO)algorithm and extreme learning machine(ELM)is proposed.Using fuzzy c-means clustering(FCM)and grey correlation analysis(GRA)to screen out key temperature measurement points of the feed system;Then,the snake optimization algorithm is used to optimize the input layer weights and hidden layer biases of the limit learning machine,and the SO-ELM thermal error prediction model is constructed using the temperature rise data and thermal error data of key temperature measurement points.To verify the accuracy and applicability of the model,a comparative analysis was conducted with the thermal error prediction models based on SSA-BP and LSMT.The results showed that the root mean square error and determination coefficient of the SO-ELM model prediction results were better than those of SSA-BP and LSTM models,which can better predict the thermal error of the machine tool feed system and provide a new idea for the compensation of machine tool thermal error prediction.