Flatness Prediction of Cold Rolled Strip Based on Deep Neural Network with Improved Activation Function

With the improvement of industrial requirements for the quality of cold rolled strips, flatness has become one of the most important indicators for measuring the quality of cold rolled strips. In this paper, the strip production data of a 1250 mm tandem cold mill in a steel plant is modeled by an improved deep neural network (the improved DNN) to improve the accuracy of strip shape prediction. Firstly, the type of activation function is analyzed, and the monotonicity of the activation function is deemed independent of the convexity of the loss function in the deep network. Regardless of whether the activation function is monotonic, the loss function is not strictly convex. Secondly, the non-convex optimization of the loss functionextended from the deep linear network to the deep nonlinear network, is discussed, and the critical point of the deep nonlinear network is identified as the global minimum point. Finally, an improved Swish activation function based on batch normalization is proposed, and its performance is evaluated on the MNIST dataset. The experimental results show that the loss of an improved Swish function is lower than that of other activation functions. The prediction accuracy of a deep neural network (DNN) with an improved Swish function is 0.38% more than that of a deep neural network (DNN) with a regular Swish function. For the DNN with the improved Swish function, the mean square error of the prediction for the flatness of cold rolled strip is reduced to 65% of the regular DNN. The accuracy of the improved DNN is up to and higher than the industrial requirements. The shape prediction of the improved DNN will assist and guide the industrial production process, reducing the scrap yield and industrial cost.

SHAPR - An AI approach to predict 3D cell shapes from 2D microscopic images

Reconstruction of shapes, forms, and sizes of three-dimensional (3D) objects from two-dimensional (2D) information is one of the most complex functions of the human brain. It also poses an algorithmic challenge and at present is a widely studied subject in computer vision. We here focus on the single cell level and present a neural network-based SHApe PRediction autoencoder SHAPR that accurately reconstructs 3D cellular and nuclear shapes from 2D microscopic images and may have great potential for application in the biomedical sciences.

Towards Self-shaping Metamaterial Shells:

Double curvature enables elegant and material-efficient shell structures, but their construction typically relies on heavy machining, manual labor, and the additional use of material wasted as one-off formwork. Using a material’s intrinsic properties for self-shaping is an energy and resource-efficient solution to this problem. This research presents a fabrication approach for self-shaping double-curved shell structures combining the hygroscopic shape-changing and scalability of wood actuators with the tunability of 3D-printed metamaterial patterning. Using hybrid robotic fabrication, components are additively manufactured flat and self-shape to a pre-programmed configuration through drying. A computational design workflow including a lattice and shell-based finite element model was developed for the design of the metamaterial pattern, actuator layout, and shape prediction. The workflow was tested through physical prototypes at centimeter and meter scales. The results show an architectural scale proof of concept for self-shaping double-curved shell structures as a resource-efficient physical form generation method.

Shape Prediction of the Sheet in Continuous Roll Forming Based on the Analysis of Exit Velocity

Continuous roll forming (CRF) is a new technology that combines continuous forming and multi-point forming to produce three-dimensional (3D) curved surfaces. Compared with other methods, the equipment of CRF is very simple, including only a pair of bendable work rolls and the corresponding shape adjustment and support assembly. By controlling the bending shapes of the upper and lower rolls and the size of the roll gap during forming, double curvature surfaces with different shapes can be produced. In this paper, a simplified expression of the exit velocity of the sheet is provided, and the formulas for the calculation of the longitudinal curvature radius are further derived. The reason for the discrepancy between the actual and predicted values of the longitudinal radius is deeply discussed from the perspective of the distribution of the exit velocity. By using the response surface methodology, the effects of the maximum compression ratio, the sheet width, the sheet thickness, and the transverse curvature radius on the longitudinal curvature radius are analyzed. Meanwhile, the correction coefficients of the predicted formulas for the positive and negative Gaussian curvature surfaces are obtained as 1.138 and 0.905, respectively. The validity and practicability of the modified formulas are verified by numerical simulations and forming experiments.

The Machine Learning's Classification Methods Comparison to Estimate Electrofacies Type, Lithology and Hydrocarbon Fluids from Geophysical Well Log Data

Supervised learning methods from machine learning are starting to be widely used in oil & gas data management. The usage of the method is adjusted to the purpose of data processing, including data classification and regression. In this research, there are six classification methods to estimate the electrofacies shape, lithology type, and fluids, namely Decision Tree (DT), Gaussian Naïve Bayes (GNB), K-Nearest Neighbors (KNN), Random Forest (RF), Support Vector Machine (SVM), and Gradient Boosting (XGB). This research compared those six methods qualitatively and quantitatively to obtain the best method. This research was conducted in the Maju Royal Field using one oil well data for training data and another one well as testing data. For validation purposes, 85% of the data was split for training and 15% for validation, aiming to evaluate the machine learning model through the correlation coefficient value. In the test data, qualitative and quantitative analyzes were also conducted. Qualitative analysis was performed by comparing the results of the electrofacies shape prediction with the original interpretation, lithology prediction with shale volume data, and prognosis of fluids with test zone data. Meanwhile, quantitatively, it is done by comparing the correct predictive data with the actual amount of data on each parameter. The training data evaluation result shows that KNN and XGB are suitable for electrofacies shape prediction. Meanwhile, lithology and fluid estimation are good with DT, KNN, and XGB methods. The qualitative and quantitative analysis result from the test data shows that the DT and GNB methods are suitable for estimating the electrofacies shape. In contrast, all methods are considered good at predicting and have good correlation values for calculating the lithology and fluids. Hence, both training and test data evaluation result has good correlation values

Adjustment of parameters of precision stamping of gear wheels on results of modeling of physical processes at their chemical and thermal processing

The surface layer after chemical-thermal treatment in structure and physico-mechanical properties differs sharply from the inner layers of the product, which leads to significant internal stresses that cause deformation and warping, i.e. resizing and shape. Prediction of the phase composition, depth of carbon saturation of the layer, microhardness and deformation of the surface of the product elements after chemical-thermal treatment based on modeling in the application package allows even at the stage of technological preparation for the production of precision hot die forging to make adjustments to the geometry of the die tool to increase dimensional accuracy, and accordingly, the durability of the gear ring gears. The input data for modeling the processes of chemical-thermal treatment is a 3D model of the product with a finite element grid and the fixation scheme of the product, temperature and time modes (heating temperature, heating and cooling rate, holding time, number of cycles), type of carburizing cooling medium and its temperature, material details.