This study investigates the application of regression neural networks, particularly the fitrnet model, in predicting the hardness of steels. The experiments involve extensive tuning of hyperparameters using Bayesian optimization and employ 5-fold and 10-fold cross-validation schemes. The trained models are rigorously evaluated, and their performances are compared using various metrics, such as mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). The results provide valuable insights into the models’ effectiveness and their ability to generalize to unseen data. In particular, Model 4208 (8-85-141-1) emerges as the top performer with an impressive RMSE of 1.0790 and an R² of 0.9900. The model, which was trained with different datasets for nearly 40 steel grades, enables the prediction of hardenability curves, but is limited to the range of the training dataset. The research paper contains an illustrative example that demonstrates the practical application of the developed model in determining the hardenability band for a specific steel grade and shows the effectiveness of the model in predicting and optimizing heat treatment results.

Despite Fused Deposition Modeling (FDM) being an economical 3D printing method known for its material versatility and ease of use, the mechanical performance of FDM-produced components is significantly influenced by process parameter settings. This study investigated the effects of the layer thickness, raster angle, build orientation, and extrusion temperature on the ultimate tensile strength (UTS) and elastic modulus of Polylactic Acid (PLA) specimens using Taguchi methods, with significance analyzed through analysis of variance (ANOVA). The results indicated that the build orientation is the primary factor affecting both the UTS and elastic modulus, with a flat orientation yielding the best performance. ANOVA showed that the build orientation, raster angle, and extrusion temperature significantly influence the UTS, with the build orientation contributing 98.16%. For the elastic modulus, the build orientation and raster angle were significant, contributing 94.83% and 1.76%, respectively. The optimal parameters were a 0.16 mm layer thickness, flat build orientation, 30°/−60° raster angle, and 200 °C extrusion temperature, resulting in predicted UTS and elastic modulus values with error percentages of 4.33% and 2.74%, respectively, compared to experimental values. The regression model demonstrated high predictive accuracy, with R-squared values of 99.71% for the UTS and 99.52% for the elastic modulus.

Artificial neural networks are an effective and frequently used modelling method in regression and classification tasks in the area of steels and metal alloys. New publications show examples of the use of artificial neural networks in this area, which appear regularly. The paper presents an overview of these publications. Attention was paid to critical issues related to the design of artificial neural networks. There have been presented our suggestions regarding the individual stages of creating and evaluating neural models. Among other things, attention was paid to the vital role of the dataset, which is used to train and test the neural network and its relationship to the artificial neural network topology. Examples of approaches to designing neural networks by other researchers in this area are presented.

In dye-sensitized solar cells (DSSCs), materials classified as Transparent Conducting Oxides (TCOs) have the capacity to conduct electricity and transmit light at the same time. Their exceptional blend of optical transparency and electrical conductivity makes them popular choices for transparent electrodes in DSSCs. Fluorine Tin Oxide (FTO) was utilized in this experiment. The optical and electrical characteristics of TCOs may be negatively impacted by their frequent exposure to hostile environments and potential for deterioration. TCOs are coated with passivating layers to increase their performance, stability, and defense against environmental elements including oxygen, moisture, and chemical pollutants. Because of its superior dielectric qualities, strong chemical stability, and suitability with TCO materials, aluminum oxide (Al₂O₃) was utilized as a passivating layer for the FTO. In this research work, Al₂O₃ was deposited via atomic layer deposition (ALD) to form thin mesoporous layers as a passivator in the photoanode (working electrode). The work focuses on finding an appropriate thickness of Al₂O₃ for optimum performance of the dye-sensitized solar cells. The solar simulation and sheet resistance analysis clearly showed 200 cycles of Al₂O₃ to exhibit an efficiency of 4.31%, which was the most efficient performance. The surface morphology and topography of all samples were discussed and analyzed.

The article presents a hybrid method for calculating the chemical composition of steel with the required hardness after cooling from the austenitizing temperature. Artificial neural networks (ANNs) and genetic algorithms (GAs) were used to develop the model. Based on 550 diagrams of continuous cooling transformation (CCT) of structural steels available in the literature, a dataset of experimental data was created. Artificial neural networks were used to develop a hardness model describing the relationship between the chemical composition of the steel, the austenitizing temperature, and the hardness of the steel after cooling. A genetic algorithm was used to identify the chemical composition of the steel with the required hardness. The value of the objective function was calculated using the neural network model. The developed method for identifying the chemical composition was implemented in a computer application. Examples of calculations of mass concentrations of steel elements with the required hardness after cooling from the austenitizing temperature are presented. The model proposed in this study can be a valuable tool to support chemical composition design by reducing the number of experiments and minimizing research costs.