基于热图像的GWOA-BiLSTM机床主轴热误差预测

Thermal error prediction of machine tool spindle of GWOA-BiLSTM based on thermal image

热误差是影响高精密数控机床加工精度的重要因素。为了提高机床加工精度和性能,减少机床运行中产生的热误差,文章提出一种基于热图像的灰狼优化算法(grey wolf optimization algorithm,GWOA)和双向长短期记忆神经网络(bidirectional long short-term memory,BiLSTM)混合的热误差预测模型。首先,采用热成像仪获取机床主轴区域的温度场信息;其次,利用DBSCAN聚类(density-based spatial clustering of applications with noise)算法和相关系数法筛选出温度敏感点;然后,通过模拟灰狼群体捕食行为,在参数空间中进行搜索以找到BiLSTM所需的最优参数;最后,使用获得的机床温度敏感点和热位移数据进行热误差预测,并在试验机床上进行验证。实验结果表明,使用GWOA优化BiLSTM神经网络的预测模型相比BiLSTM神经网络预测模型的均方根误差(root mean square error,RMSE)和平均绝对误差(mean absolute error,MAE)分别减小了约0.5180、0.3823μm,决定系数R^(2)提升了0.0578。与BiLSTM神经网络模型相比,利用GWOA优化后的模型具有更加优良的预测性能。

Thermal error is an important factor affecting the machining accuracy of high precision CNC machine tools.In order to improve the machining accuracy and performance of the machine tool and reduce the thermal error generated during the operation of the machine tool,a grey wolf optimization algorithm(GWOA)and bidirectional long short-term memory(BiLSTM)neural network based on thermal images are proposed in our study.Firstly,the thermal imager is used to obtain the temperature field information of the spindle area of the machine tool.Secondly,the temperature sensitive points were screened by DBSCAN algorithm and correlation coefficient method.Then,by simulating the predation behavior of gray wolves,the optimal parameters of BiLSTM were found by searching in the parameter space.Finally,the thermal error is predicted by using the temperature-sensitive points and thermal displacement data obtained,and verified on the test machine.The experimental results show that compared with the root mean square error(RMSE)and mean absolute error(MAE)of the BiLSTM neural network prediction model optimized by GWOA,MAE decreased by 0.5180 and 0.3823μm,2 respectively,and the coefficient of determination R^(2) increased by 0.0578.Compared with BiLSTM neural network model,GWOA optimized model has better prediction performance.