Numerical Control Milling Energy Consumption Prediction Method Considering Incomplete Data

doi:10.3901/JME.2025.19.341

Abstract

Abstract: In the process of CNC milling, the collected data is incomplete due to equipment failure and improper operation, which leads to the problem of low accuracy of CNC milling energy consumption prediction. By analyzing the energy consumption characteristics of CNC milling and the circumstances associated with missing data, a prediction method of CNC milling energy consumption considering data incompleteness is proposed. Firstly, the missing values are estimated based on the denoising auto-encoder expectation maximum multiple interpolation (EMMIDA) to generate new complete data. Secondly, The FEDformer algorithm is used to construct the energy consumption prediction model, and the energy consumption prediction of CNC milling is realized. Finally, the proposed method is compared with other common data filling method and prediction algorithms at different miss rates. Experimental results show that even if the miss rate increases to 30%, the MAE of the EMMIDA filling method is reduced by 20.3%, 25.9% and 44.6% respectively compared with multiple imputation by chained equations (MICE), generative adversarial imputation networks (GAIN), and K-nearest neighbors (KNN), and the RMSE is reduced by 17.8%, 36.5% and 42.1% respectively compared with the Transformer. The MAE and RMSE of the FEDformer energy consumption prediction model are reduced by 4.36% and 5.74% compared with the Transformer, and the R² is increased by 7.73%, which are better than Informer, LTSM and CNN, the proposed method can effectively reduce the prediction error caused by the missing energy consumption data of CNC milling.

Key words: CNC milling, energy consumption prediction, missing data, EMMIDA imputation, FEDformer

CLC Number:

TH547

ZHAO Jixuan, LI Congbo, WU Wei, ZHNAG You, JING Peifeng. Numerical Control Milling Energy Consumption Prediction Method Considering Incomplete Data[J]. Journal of Mechanical Engineering, 2025, 61(19): 341-351.

References

[1] CAI W，LIU F，ZHOU X N，et al. Fine energy consumption allowance of workpieces in the mechanical manufacturing industry[J]. Energy，2016，114：623-633.
[2] CHEN X Z，LI C B，TANG Y，et al. Energy efficient cutting parameter optimization[J]. Frontiers of Mechanical Engineering，2021，16(2)：221-248.
[3] DIAZ N，NINOMIYA K，NOBLE J，et al. Environmental impact characterization of milling and implications for potential energy savings in industry[J]. Procedia CIRP，2012，1：518-523.
[4] HE Y，LIU F，WU T，et al. Analysis and estimation of energy consumption for numerical control machining[J]. Proceedings of the Institution of Mechanical Engineers，Part B：Journal of Engineering Manufacture，2012，226(2)：255-266.
[5] LI C，XIAO Q，TANG Y，et al. A method integrating Taguchi，RSM and MOPSO to CNC machining parameters optimization for energy saving[J]. Journal of Cleaner Production，2016，135：263-275.
[6] LV Y，LI C，HE J，et al. Energy saving design of the machining unit of hobbing machine tool with integrated optimization[J]. Frontiers of Mechanical Engineering，2022，17(3)：38.
[7] DIETRICH B，WALTHER J，WEIGOLD M，et al. Machine learning based very short term load forecasting of machine tools[J]. Applied Energy，2020，276：115440.
[8] 易茜，罗雨松，胡春晖，等. 共享制造下基于数据迁移的滚齿加工碳耗预测方法[J]. 中国机械工程，2023，34(15)：1820-1831. YI Qian，LUO Yusong，HU Chunhui，et al. Carbon consumption prediction method of gear hobbing based on data migration in shared manufacturing[J]. China Mechanical Engineering，2023，34(15)：1820-1831.
[9] 李聪波，尹誉先，肖溱鸽，等. 数据驱动下基于元动作的数控车削能耗预测方法[J]. 中国机械工程，2020，31(21)：2601-2611. LI Congbo，YIN Yuxian，XIAO Qinge，et al. Data-driven energy consumption prediction method of CNC turning based on meta-action[J]. China Mechanical Engineering，2020，31(21)：2601-2611.
[10] LI W，LI B，HE S，et al. A novel milling parameter optimization method based on improved deep reinforcement learning considering machining cost[J]. Journal of Manufacturing Processes，2022，84：1362-1375.
[11] GUO Z，WAN Y，YE H. A data imputation method for multivariate time series based on generative adversarial network[J]. Neurocomputing，2019，360：185-197.
[12] DEMPSTER A P，LAIRD N M，RUBIN D B. Maximum likelihood from incomplete data via the EM algorithm[J]. Journal of the Royal Statistical Society，1977，39(1)：1-22.
[13] WANG Z，WANG L，TAN Y，et al. Fault detection based on bayesian network and missing data imputation for building energy systems[J]. Applied Thermal Engineering，2021，182：116051.
[14] LEE W，LEE S，BYUN J，et al. Variational cycle- consistent imputation adversarial networks for general missing patterns[J]. Pattern Recognition，2022，129：108720.
[15] 张友，李聪波，林利红，等. 数据不完备下基于Informer的离心鼓风机故障趋势预测方法[J]. 计算机集成制造系统，2023，29(1)：133-145. ZHANG You，LI Congbo，LIN Lihong，et al. Centrifugal blower fault trend prediction method based on Informer with incomplete data[J]. Computer Integrated Manufacturing System，2023，29(1)：133-145.
[16] PAN J，LI C，TANG Y，et al. Energy consumption prediction of a CNC machining process with incomplete data[J]. IEEE/CAA Journal of Automatica Sinica，2021，8(5)：987-1000.
[17] 易茜，柳淳，李聪波等. 数据缺失下基于改进生成对抗填补网络的碳耗预测方法[J]. 机械工程学报，2023，59(11)：264-275. YI Qian，LIU Chun，LI Congbo，et al. Prediction method of hobbing carbon consumption based on improved generative adversarial imputation net with missing data[J]. Journal of Mechanical Engineering，59(11)：264-275.
[18] KWAK S K，KIM J H. Statistical data preparation：management of missing values and outliers[J]. Korean Journal of Anesthesiology，2017，70(4)：407-411.
[19] AZUR M J，STUART E A，FRANGAKIS C，et al. Multiple imputation by chained equations：What is it and how does it work?[J]. International Journal of Methods in Psychiatric Research，2011，20(1)：40-49.
[20] ZHOU T，MA Z，WEN Q，et al. Fedformer：Frequency enhanced decomposed transformer for long-term series forecasting[C]//International Conference on Machine Learning. PMLR，2022：27268-27286.
[21] SUN Y，LI J，XU Y，et al. Deep learning versus conventional methods for missing data imputation：A review and comparative study[J]. Expert Systems with Applications，2023：120201.