Study of static thermal deformation modeling based on a hybrid CNN-LSTM model with spatiotemporal correlation

Guo, Jiahao; Xiong, Qingyu; Chen, Jing; Miao, Enming; Wu, Chao; Zhu, Qiwu; Yang, Zhengyi; Chen, Jie

doi:10.1007/s00170-021-08462-9

Cited by 27 publications

(8 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Guo et al [ 23 ] study the spatiotemporal correlation of data in the static thermal deformation modeling of CNC machine tools. The methods use a hybrid CNN-LSTM model with spatiotemporal correlation (ST-CLSTM).…”

Section: Background Review and Related Workmentioning

confidence: 99%

An Intelligent Thermal Compensation System Using Edge Computing for Machine Tools

Kristiani,

Wang,

Liu

et al. 2024

Sensors

View full text Add to dashboard Cite

This paper focuses on the use of smart manufacturing in lathe-cutting tool machines, which can experience thermal deformation during long-term processing, leading to displacement errors in the cutting head and damage to the final product. This study uses time-series thermal compensation to develop a predictive system for thermal displacement in machine tools, which is applicable in the industry using edge computing technology. Two experiments were carried out to optimize the temperature prediction models and predict the displacement of five axes at the temperature points. First, an examination is conducted to determine possible variances in time-series data. This analysis is based on the data obtained for the changes in time, speed, torque, and temperature at various locations of the machine tool. Using the viable machine-learning models determined, the study then examines various cutting settings, temperature points, and machine speeds to forecast the future five-axis displacement. Second, to verify the precision of the models created in the initial phase, other time-series models are examined and trained in the subsequent phase, and their effectiveness is compared to the models acquired in the first phase. This work also included training seven models of WNN, LSTNet, TPA-LSTM, XGBoost, BiLSTM, CNN, and GA-LSTM. The study found that the GA-LSTM model outperforms the other three best models of the LSTM, GRU, and XGBoost models with an average precision greater than 90%. Based on the analysis of training time and model precision, the study concluded that a system using LSTM, GRU, and XGBoost should be designed and applied for thermal compensation using edge devices such as the Raspberry Pi.

show abstract

Section: Background Review and Related Workmentioning

confidence: 99%

An Intelligent Thermal Compensation System Using Edge Computing for Machine Tools

Kristiani,

Wang,

Liu

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…To reduce the length of this work, the theoretical descriptions of the models used in this paper are not presented because they are well explained elsewhere in different sources. The detail of the CLSTM theory has been introduced in [14][15][16][17] for successfully predicting various datasets including solar photosynthetic photon ux density, solar radiation, photovoltaic power, and thermal displacement. Additionally, the studies in [18][19][20][21][22][23] present the mathematical formulas of the other deep learning models (LSTM, CNN, DNN, MLP) used in this study.…”

Section: Theoretical Overviewmentioning

confidence: 99%

Multi-step solar UV index prediction using deep learning methods

AL‐Musaylh

Ghimire

Al‐Daffaie

et al. 2023

Preprint

View full text Add to dashboard Cite

The impact of ultraviolet (UV) radiation on public health is severe and can cause sunburn, skin aging and cancer, immunosuppression, and eye damage. Minimization of exposure to solar UV is required in order to reduce the risks of these illnesses to the public. Greater public awareness and a prediction of the ultraviolet index (UVI) is considered an essential task for the minimization of solar UV exposures. This research has designed artificial intelligence based deep learning models to predict multistep UVI index. It has developed a convolutional neural network integrated with long short-term memory network (CLSTM) as the main model to forecast UVI for Brisbane with latitude − 27.47 and longitude 153.02, the capital city of Queensland, Australia. Solar zenith angle (SZA) data were used together with UVI as inputs in the CLSTM for 10-min, 30-min and 60-min UVI prediction. The CLSTM model was benchmarked against long short-term memory network (LSTM), convolutional neural network (CNN), Deep Neural Network (DNN), multilayer perceptron (MLP), extreme learning machine (ELM), random forest regression (RFR), Extreme Gradient Boosting (XGB), and Pro6UV Deterministic models. The experimental results showed that the CLSTM model outperformed these models with RMSE = 0.3817, MAE = 0.1887, RRMSE = 8.0086%, MAPE = 4.6172% and APB = 3.9586 for 10-min prediction. In addition to that, these metrics for 30-min and 60-min prediction were RMSE = 0.4866/0.5146, MAE = 0.2763/0.3038, RRMSE = 10.4860%/11.5840%, MAPE = 8.1037%/9.6558% and APB = 5.9546/6.8386, respectively. Thus, the CLSTM model can yield improved UVI prediction for both the public and the government agencies.

show abstract

“…However, the LSTM is prone to the problems of vanishing gradients and exploding gradients when dealing with long sequences, which can affect the performance and training effects [15]. To simultaneously extract spatial and temporal features of signals, a new deep learning model, called CNN-LSTM is developed, and it is effective to detect mechanical faults [16]. Han et al [17] proposed a multi-scale dilated CNN-LSTM model for diagnosing unbalanced sample faults in planetary gearboxes under noisy environments, and it has higher recognition rates compared with various CNN models.…”

Section: Introductionmentioning

confidence: 99%

Rotating machinery fault classification based on one-dimensional residual network with attention mechanism and bidirectional gated recurrent unit

Dong,

Zhao,

Cui

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Conventional convolutional neural networks (CNNs) predominantly emphasize spatial features of signals and often fall short in prioritizing sequential features. As the number of layers increases, they are prone to issues such as vanishing or exploding gradients, leading to training instability and subsequent erratic fluctuations in loss values and recognition rates. To address this issue, a novel hybrid model, termed one-dimensional residual network with attention mechanism and bidirectional gated recurrent unit (1D-RAM-BGRU) is developed for rotating machinery fault classification. First, a novel one-dimensional residual network (1D-ResNet) with optimized structure is constructed to obtain spatial features and mitigate the gradient vanishing or exploding. Second, the attention mechanism (AM)is designed to catch important impact characteristics for fault samples. Next, temporal features are mined through the bidirectional gated recurrent unit (BGRU .) Finally, feature information is summarized through global average pooling, and the fully connected layer is utilized to output the final classification result for rotating machinery fault diagnosis. The developed technique which is tested on one set of planetary gear data and three different sets of bearing data, has achieved classification accuracy of 98.5%, 100%, 100%, and 100%, respectively. Compared with other methods, including CNN, CNN-BGRU, CNN-AM, and CAM-BGRU, the proposed technique has the highest recognition rate and stable diagnostic performance.

show abstract

Study of static thermal deformation modeling based on a hybrid CNN-LSTM model with spatiotemporal correlation

Cited by 27 publications

References 27 publications

An Intelligent Thermal Compensation System Using Edge Computing for Machine Tools

An Intelligent Thermal Compensation System Using Edge Computing for Machine Tools

Multi-step solar UV index prediction using deep learning methods

Rotating machinery fault classification based on one-dimensional residual network with attention mechanism and bidirectional gated recurrent unit

Contact Info

Product

Resources

About