Abstract:This paper presents the development and evaluation of neural network models using a small input–output dataset to predict the thermal behavior of a high-speed motorized spindles. Different neural multi-output regression models were developed and evaluated using Keras, one of the most popular deep learning frameworks at the moment. ANN was developed and evaluated considering the following: the influence of the topology (number of hidden layers and neurons within), the learning parameter, and validation techniqu… Show more
“…where Q(t) is the time-varying function of heat flow from the heating component, and V is the volume of the heat source component. The heating of the motor and bearings is the main cause of temperature rise in the spindle system [18,19]. Due to the installation of the motor outside the spindle box, less heat is transferred to the spindle components through the mounting surface of the box.…”
Section: Time-varying Factors Of System Heat Sourcementioning
The thermal characteristics of the spindle system for CNC machine tools are influenced by multiple factors which are nonlinear and time-varying. In this paper, a nonlinear time-varying thermal characteristics solving model for the spindle system was established based on the numerical solution method. Through theoretical deduction and data fitting, mathematical models of nonlinear time-varying factors including the friction torque generated by lubricants, convective heat transfer coefficient, and coolant and ambient temperature are constructed. The temperature and displacement of the spindle system at each time step are solved by considering the comprehensive effect of multiple nonlinear time-varying factors. And the actual temperature and axial deformation data of the spindle system are obtained through thermal characteristics detection experiments. By comparing solution results affected by multiple nonlinear time-varying factors and by non time-varying factors with experimental data, it can be concluded that the nonlinear time-varying thermal characteristics model has advantages in reflecting the trend of numerical changes and the accuracy of result solving over a method considering non time-varying factors and the solution values of temperature affected by multiple nonlinear time-varying factors are almost consistent with detection values and the relative errors are all within ±3%. The relative error of axial deformation between the value solved by the model and the detection value is close to −1%. This conclusion demonstrates the rationality and accuracy of the thermal characteristics solving model and the construction of nonlinear time-varying factors. This study is of great significance for exploring the thermal characteristics of the spindle system and improving CNC machine tool performance in depth.
“…where Q(t) is the time-varying function of heat flow from the heating component, and V is the volume of the heat source component. The heating of the motor and bearings is the main cause of temperature rise in the spindle system [18,19]. Due to the installation of the motor outside the spindle box, less heat is transferred to the spindle components through the mounting surface of the box.…”
Section: Time-varying Factors Of System Heat Sourcementioning
The thermal characteristics of the spindle system for CNC machine tools are influenced by multiple factors which are nonlinear and time-varying. In this paper, a nonlinear time-varying thermal characteristics solving model for the spindle system was established based on the numerical solution method. Through theoretical deduction and data fitting, mathematical models of nonlinear time-varying factors including the friction torque generated by lubricants, convective heat transfer coefficient, and coolant and ambient temperature are constructed. The temperature and displacement of the spindle system at each time step are solved by considering the comprehensive effect of multiple nonlinear time-varying factors. And the actual temperature and axial deformation data of the spindle system are obtained through thermal characteristics detection experiments. By comparing solution results affected by multiple nonlinear time-varying factors and by non time-varying factors with experimental data, it can be concluded that the nonlinear time-varying thermal characteristics model has advantages in reflecting the trend of numerical changes and the accuracy of result solving over a method considering non time-varying factors and the solution values of temperature affected by multiple nonlinear time-varying factors are almost consistent with detection values and the relative errors are all within ±3%. The relative error of axial deformation between the value solved by the model and the detection value is close to −1%. This conclusion demonstrates the rationality and accuracy of the thermal characteristics solving model and the construction of nonlinear time-varying factors. This study is of great significance for exploring the thermal characteristics of the spindle system and improving CNC machine tool performance in depth.
“…Several suthors [23][24][25][26][27][28][29] employed neural networks featuring one or two hidden layers to predict surface roughness using a relatively small dataset of 27 samples. They experimented by adjusting the number of hidden layers, varying the number of neurons within those layers, and employing different training algorithms to identify the neural network configuration that yielded the best performance.…”
Section: Literature Reviewmentioning
confidence: 99%
“…Some researchers, such as Kosarac et al [27], leveraged neural networks and relatively small datasets to predict phenomena beyond surface roughness. They employed an Artificial Neural Network (ANN) model to analyze how working conditions influence the thermal behavior of a motorized grinder spindle.…”
Mechanical engineering plays an important role in the design and manufacture of medical devices, implants, prostheses, and other medical equipment, where the machining of bio-compatible materials have a special place. There are a lot of different conventional and non-conventional types of machining of biocompatible materials. One of the most frequently used methods is milling. The first part of this research explores the machining parameters optimization minimizing surface roughness in milling titanium alloy Ti-6Al-4V. A full factorial design involving four factors (cutting speed, feed rate, depth of cut, and the cooling/lubricating method), each having three levels, implies the 81 experimental runs. Using the Taguchi method, the number of experimental runs was reduced from 81 to 27 through an orthogonal design. According to the analysis of variance (ANOVA), the most significant parameter for surface roughness is feed rate. The second part explores the possibilities of using different ML techniques to create a predictive model for average surface roughness using the previously created small datasets. The paper presents a comparative analysis of several commonly used techniques for handling small datasets and regression problems. The best results indicate that the widely used machine learning algorithm Random Forest excels in handling regression problems and small datasets.
“…In recent years, artificial intelligence has gradually emerged, underpinning systems such as the BP neural network, RBF neural network and SVM, which have gradually been applied to predict the hazard of water and mud inrush disasters in tunnels [7][8][9][10][11][12]. Although these methods are all used to determine the hazard of water inrush in tunnels, the performances of the models are different.…”
To prevent large-scale water inrush accidents during the excavation process of a water-rich tunnel, a method, based on a random forest (RF) algorithm, for predicting the hazard level of water inrush is proposed. By analyzing hydrogeological conditions, six factors were selected as evaluating indicators, including stratigraphic lithology, inadequate geology, rock dip angle, negative terrain area ratio, surrounding rock grade, and hydrodynamic zonation. Through the statistical analysis of 232 accident sections, a dataset of water inrush accidents in water-rich tunnels was established. We preprocessed the dataset by detecting and replacing outliers, supplementing missing values, and standardizing the data. Using the RF model in machine learning, an intelligent prediction model for the hazard of water inrush in water-rich tunnels was established through the application of datasets and parameter optimization processing. At the same time, a support vector machine (SVM) model was selected for comparison and verification, and the prediction accuracy of the RF model reached 98%, which is higher than the 87% of the SVM. Finally, the model was validated by taking the water inrush accident in the Yuanliangshan tunnel as an example, and the predicted results have a high degree of consistency with the actual hazard level. This indicates that the RF model has good performance when predicting water inrush in water-rich tunnels and that it can provide a new means by which to predict the hazard of water inrush in water-rich tunnels.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.