Optimization Procedure Research Articles

Micropropagation and cryopreservation of the rare endemic Colchicum figlalii germplasm

BACKGROUND: The natural population of Colchicum figlalii (Varol) Parolly & Eren grows in a narrow area of serpentine rock clearings at an altitude of 1900-2100 m in Southwestern Anatolia (Sandras Mountain, Mugla, Turkey). The species is regarded as endangered according to the IUCN Red List Categories. OBJECTIVE: To develop an optimum procedure for in vitro propagation and cryopreservation of germplasm of this rare endemic. MATERIALS AND METHODS: A total of 281 bulbs were used as in vitro culture starting material and after surface sterilization, clean material was obtained from 157 of them. Woody Plant Medium (WPM), Olive Medium (OM), and Murashige and Skoog medium (MS) were used for in vitro culture establishment. RESULTS: The maximum regeneration rate (˜67.3%) was obtained after four weeks of incubation on OM. The calli were successfully induced by using OM supplemented with 10.7 μM NAA from leaves of in vitro grown C. figlalii bulbs. A PVS2-vitrification procedure was used for cryopreservation of C. figlalii callus tissue. After cryo-storage, the best result for regeneration (66.7%) was obtained from calli treated with PVS2 for 75 min before plunging into liquid nitrogen. All rooted seedlings derived from cryopreserved calli were successfully acclimatized to greenhouse conditions. CONCLUSION: This study is an effective reference for future long-term conservation of similar species that are difficult to cryopreserve.

Geometry Optimization of Industry Buildings for Energy-Driven Design Development in Bangladesh

Even though the manufacturing industry consumes roughly 54% of total available energy globally, little consideration has been devoted to optimizing energy in the early stages of industry design, particularly in densely populated cities. With the increased demand for green buildings, energy performance has a greater influence on design results. As a result, this paper provides an envelope optimization technique that can assist architects and computational designers in analyzing the environmental performance of various alternatives and developing optimal design solutions, especially for Bangladesh and the building of these regions. First, an existing industrial site in Dhaka, Bangladesh's capital was chosen as a case study, and a hypothetical industry building, as well as its surroundings, were parametrically developed. Then, for the optimization method, the design factors linked to the building envelopes were chosen. Finally, a Multi-objective Optimization (MOO) procedure was utilized for defining performance measures including daylighting, energy and comfort measures, UDI, EUI, and PPD. According to the MOO results, the UDI may be enhanced by 25.286% as compared to the least favorable scenario. Consequently, the EUI may decreased by 38.718 kWh/m2 while the PPD can be increased by 41.78%. The geometric configuration of East, West, North, and South played a significant role when designing the industrial building. According to the analysis, the geometric configuration of a South WWR of 50%, a West WWR of 30%, an East WWR of 70%, and a sill height of 0.75m is the most feasible option. A statistical analysis of design factors and performance measures demonstrated that the window-to-wall ratio, particularly on south walls, has the greatest impact on industrial building design in densely populated areas. The proposed approach is expected to be used by architects and municipal planners to develop design metrics based on simulation results.

Short-term effects of gastric bypass versus sleeve gastrectomy on high LDL cholesterol: The BASALTO randomized clinical trial

BackgroundThere has been a substantial increase in the use of laparoscopic sleeve gastrectomy (SG) to treat morbid obesity despite observational evidence demonstrating the superiority of Roux-en-Y gastric bypass (RYGB) for reducing low-density lipoprotein (LDL) cholesterol. The main aim was to ascertain whether high LDL cholesterol levels should be considered when selecting the most appropriate surgical procedure for each patient (RYGB or SG).MethodsIn this single-center, randomized clinical trial using intention-to-treat analysis, 38 patients with severe obesity and elevated levels of LDL cholesterol were randomly assigned to undergo RYGB or SG. The primary outcome was LDL cholesterol remission at 12 months, defined as LDL cholesterol < 3.36 nmol/l without lipid-lowering medications. Secondary outcomes included changes in weight, other comorbidities, qualitative lipoprotein traits, cholesterol esters, glycoproteins, cholesterol absorption and synthesis metabolites and complications.ResultsIntention-to-treat analysis revealed that LDL cholesterol remission occurred in 66.6% of RYGB patients compared to 27.8% of SG patients (p = 0.019). Among patients completing follow-up, RYGB demonstrated superior remission (80.0% vs. 29.4%, p = 0.005). Exclusive benefits of RYGB included a reduction in large, medium, and small LDL particles. Cholesterol absorption markers showed differential behavior after both techniques: campesterol (Δ −15.2 µg/mg, 95% CI −30.2 to −0.1) decreased after RYGB, and sitosterol (Δ 21.1 µg/mg, 95% CI 0.9 to 41.2), cholestanol (Δ 30.6 µg/mg, 95% CI 14.8 to 57.9) and campesterol (Δ 18.4 µg/mg, 95% CI 4.4 to 32.3) increased after SG. No differences in weight loss, cholesterol esters, glycoproteins, cholesterol synthesis metabolites or postoperative complications were observed between techniques.ConclusionIn conclusion, RYGB is superior to SG in terms of short-term of high LDL cholesterol remission. Furthermore, RYGB also led to a greater improvement in lipoprotein parameters that confer an atherogenic profile. Therefore, the presence of elevated levels of LDL cholesterol should be considered when determining the optimal bariatric surgery procedure for each patient.Trial registrationClinicaltrials.gov number, NCT03975478).Graphical abstract

Optimization of antimicrobial nanocomposite films based on carboxymethyl cellulose incorporating chitosan nanofibers and Guggul gum polysaccharide

The present study utilized response surface methodology (RSM) to investigate the impact of varying concentrations of carboxymethyl cellulose (CMC: 0.75–1.75 wt%), Commiphora mukul polysaccharide (CMP: 0–1 wt%), and Chitosan Nanofiber (CHNF: 0–1 wt%) on the physical and antimicrobial characteristics of nanocomposite films based on CMC. The optimization process aimed to enhance ultimate tensile strength (UTS), strain at break (SAB), and antibacterial activity, while minimizing water vapor permeability (WVP), solubility, swelling, moisture content, opacity, and total color difference (ΔE). The results revealed that both CMP and CHNF had a positive influence on reducing moisture content, WVP, and increasing UTS. However, higher concentrations of CMP and CHNF had a divergent effect on SAB, ΔE, and swelling. The incorporation of CMP led to increased opacity and solubility, while the inclusion of CHNF resulted in decreased opacity and solubility. Notably, only CHNF addition significantly improved the antibacterial properties of the films. By applying the optimization procedure utilizing RSM, the formulation containing CMC (1.5 wt%), CMP (0.25 wt%), and CHNF (0.75 wt%) demonstrated superior physical, mechanical, and antibacterial properties in the biodegradable film matrix. These findings highlight the potential of utilizing these components to enhance the performance of CMC-based nanocomposite films.

Prediction of the binding mechanism of a selective DNA methyltransferase 3A inhibitor by molecular simulation

DNA methylation is an epigenetic mechanism that introduces a methyl group at the C5 position of cytosine. This reaction is catalyzed by DNA methyltransferases (DNMTs) and is essential for the regulation of gene transcription. The DNMT1 and DNMT3A or -3B family proteins are known targets for the inhibition of DNA hypermethylation in cancer cells. A selective non-nucleoside DNMT3A inhibitor was developed that mimics S-adenosyl-l-methionine and deoxycytidine; however, the mechanism of selectivity is unclear because the inhibitor–protein complex structure determination is absent. Therefore, we performed docking and molecular dynamics simulations to predict the structure of the complex formed by the association between DNMT3A and the selective inhibitor. Our simulations, binding free energy decomposition analysis, structural isoform comparison, and residue scanning showed that Arg688 of DNMT3A is involved in the interaction with this inhibitor, as evidenced by its significant contribution to the binding free energy. The presence of Asn1192 at the corresponding residues in DNMT1 results in a loss of affinity for the inhibitor, suggesting that the interactions mediated by Arg688 in DNMT3A are essential for selectivity. Our findings can be applied in the design of DNMT-selective inhibitors and methylation-specific drug optimization procedures.

Clear-Plenoxels: Floaters free radiance fields without neural networks

Directly optimizing radiance fields with explicit 3D primitives, such as voxels and 3D Gaussians, demonstrates impressive effectiveness in both training and rendering. However, optimizing explicit 3D structures from limited training views using differentiable volumetric rendering often results in ghosting artifacts (floaters), i.e., it suffers from early-stage errors in unseen space and remains as evident ghosting artifacts as gradient becomes very sparse spatially with optimization procedure processes. Thus, to address this problem, this paper proposes Clear-Plenoxels, which involves three simple yet non-trivial components: 1) a Visual Hull-based initialization strategy for maximizing the utilization of object resolution and effectively rejecting false updates during early-stage optimization by learning rate assignment per voxel grid; 2) an effective penalty function on local grid discrepancy for ensuring consistent rendering at different view directions, which can remove unnecessary voxels; 3) a mask guided transmittance supervision for each training ray, significantly guaranteeing depth prediction precision when the object's surface color is close to the background. Experiments on a public dataset demonstrate that our method successfully overcomes ghosting artifacts by directly optimizing explicit primitive strategy. The proposed method achieves approximately 1 dB improvement in voxel grid-based methods and matches, if not surpasses, state-of-the-art quality regarding novel view rendering. The code and visualization videos are publicly available at https://github.com/nortonBryan/Clear-plenoxels.

A study of stress, composition and grain interaction gradients in energy-dispersive X-ray stress analysis on materials with cubic symmetry

The influence of various combinations of residual stress, composition and grain interaction gradients in polycrystalline materials with cubic symmetry on energy-dispersive X-ray stress analysis is theoretically investigated. For the evaluation of the simulated sin2ψ distributions, two different strategies are compared with regard to their suitability for separating the individual gradients. It is shown that the separation of depth gradients of the strain-free lattice parameter a 0(z) from residual stress gradients σ(z) is only possible if the data analysis is carried out in section planes parallel to the surface. The impact of a surface layer z* that is characterized by a direction-dependent grain interaction model in contrast to the volume of the material is quantified by comparing a ferritic and an austenitic steel, which feature different elastic anisotropy. It is shown to be of minor influence on the resulting residual stress depth profiles if the data evaluation is restricted to reflections hkl with orientation factors Γ hkl close to the model-independent orientation Γ*. Finally, a method is proposed that allows the thickness of the anisotropic surface layer z* to be estimated on the basis of an optimization procedure.

Reduced preoperative serum choline esterase levels and fecal peritoneal contamination as potential predictors for the leakage of intestinal sutures after source control in secondary peritonitis

BackgroundThe high rate of stoma placement during emergency laparotomy for secondary peritonitis is a paradigm in need of change in the current fast-track surgical setting. Despite growing evidence for the feasibility of primary bowel reconstruction in a peritonitic environment, little data substantiate a surgeons’ choice between a stoma and an anastomosis. The aim of this retrospective analysis is to identify pre- and intraoperative parameters that predict the leakage risk for enteric sutures placed during source control surgery (SCS) for secondary peritonitis.MethodsBetween January 2014 and December 2020, 497 patients underwent SCS for secondary peritonitis, of whom 187 received a primary reconstruction of the lower gastro-intestinal tract without a diverting stoma. In 47 (25.1%) patients postoperative leakage of the enteric sutures was directly confirmed during revision surgery or by computed tomography. Quantifiable predictors of intestinal suture outcome were detected by multivariate analysis.ResultsLength of intensive care, in-hospital mortality and failure of release to the initial home environment were significantly higher in patients with enteric suture leakage following SCS compared to patients with intact anastomoses (p < 0.0001, p = 0.0026 and p =0.0009, respectively). Reduced serum choline esterase (sCHE) levels and a high extent of peritonitis were identified as independent risk factors for insufficiency of enteric sutures placed during emergency laparotomy.ConclusionsA preoperative sCHE < 4.5 kU/L and generalized fecal peritonitis associate with a significantly higher incidence of enteric suture insufficiency after primary reconstruction of the lower gastro-intestinal tract in a peritonitic abdomen. These parameters may guide surgeons when choosing the optimal surgical procedure in the emergency setting.

THE METHOD OF RANKING EFFECTIVE PROJECT SOLUTIONS IN CONDITIONS OF INCOMPLETE CERTAINTY

The subject of research in the article is the process of ranking options in support systems for project decision-making under conditions of incomplete certainty. The goal of the work is to increase the efficiency of technologies for automated design of complex systems due to the development of a combined method of ranking effective options for building objects in conditions of incomplete certainty of input data. The following tasks are solved in the article: analysis of the current state of the problem of ranking options in support systems for project decision-making; decomposition of problems of system optimization of complex design objects and support of project decision-making; development of a variant ranking method that combines the procedures of lexicographic optimization and cardinal ordering in conditions of incomplete certainty of input data. The following methods are used: systems theory, utility theory, optimization, operations research, interval and fuzzy mathematics. Results. According to the results of the analysis of the problem of supporting project decision-making, the existence of the problem of correctly reducing subsets of effective options for ranking, taking into account factors that are difficult to formalize and the experience of the decision-maker (DM), was established. Decomposition of the problems of system optimization of complex design objects and support for project decision-making was carried out. For the case of ordinalistic presentation of preferences between local criteria, an estimate of the size of the rational reduction of subsets of optimal and suboptimal options for each of the indicators is proposed. Its use allows for one approach to obtain a subset of effective variants of a given capacity for analysis and final selection of the DM. A method of transforming the ordinalistic presentation of preferences between local criteria to their quantitative presentation in the form of weighting coefficients is proposed. Conclusions. The developed methods expand the methodological foundations of the automation of processes supporting the adoption of multi-criteria project decisions. They make it possible to correctly reduce the set of effective alternatives in conditions of incomplete certainty of the input data for the final choice, taking into account factors that are difficult to formalize, knowledge and experience of ODA. The practical use of the obtained results will allow to reduce the time and capacity complexity of the procedures for supporting project decision-making, and due to the use of the technology of selection of subsets of effective options with intervally specified characteristics - to guarantee the quality of project decisions and to provide a more complete assessment of them.

Pump System Model Parameter Identification Based on Experimental and Simulation Data

In this paper, the entire downhole fluid-sucker rod-pump system is replaced with a viscoelastic vibration model, namely a third-order differential equation with an inhomogeneous forcing term. Both Kelvin’s and Maxwell’s viscoelastic models can be implemented along with the dynamic behaviors of a mass point attached to the viscoelastic model. By employing the time-dependent polished rod force measured with a dynamometer as the input to the viscoelastic dynamic model, we have obtained the displacement responses, which match closely with the experimental measurements in actual operations, through an iterative process. The key discovery of this work is the feasibility of the so-called inverse optimization procedure, which can be utilized to identify the equivalent scaling factor and viscoelastic system parameters. The proposed Newton–Raphson iterative method, with some terms in the Jacobian matrix expressed with averaged rates of changes based on perturbations of up to two independent parameters, provides a feasible tool for optimization issues related to complex engineering problems with mere information of input and output data from either experiments or comprehensive simulations. The same inverse optimization procedure is also implemented to model the entire fluid delivery system of a very viscous non-Newtonian polymer modeled as a first-order ordinary differential equation (ODE) system similar to the transient entrance developing flow. The convergent parameter reproduces transient solutions that match very well with those from fully fledged computational fluid dynamics models with the required inlet volume flow rate and outlet pressure conditions.

Tailoring Heterojunctions in CsPbBrxCl3‐x – MoS2 Composites for Efficient Photocatalysis and Hydrogen Evolution

We demonstrate appropriate tuning of heterojunctions in CsPbBrxCl3‐x – MoS2 composites (where x=0,1,2,3) by controlled regulation of the halide stoichiometry in the perovskite. A thorough optimization procedure determined the most effective photocatalyst, considering the pristine MoS2, perovskites with varying halide ratios, various physical mixing ratios of the two, and in‐situ synthesized composite ratios of CsPbBrxCl3‐x and MoS2 (2:1, 1.5:1, 1:1, 1:1.5, 1:2). Under two hours of exposure to visible light, a remarkable photocatalytic performance of CsPbBrCl2:MoS2 with a 1:2 ratio was observed, removing 98% of the methylene blue (MB) dye. Notably, only the CsPbBrCl2 and MoS2 composite demonstrated higher efficiencies since it resulted in a n‐n type II heterojunction. Additionally, the CsPbBrCl2:MoS2 composite exhibits the highest reaction rate constant, fifteen times higher than the pristine perovskite. Reusability assessment of this combination revealed sustained activity of 87% for up to 5 cycles. The hydrogen evolution reaction investigations were carried out using the optimized CsPbBrCl2:MoS2 composite, which yielded 265 times more hydrogen than pristine CsPbBrCl2. The Faradaic efficiency for 1:2 CsPbBrCl2:MoS2 was found to be 96.61%. Our results offer crucial perspectives on optimizing perovskite‐MoS2 composites and demonstrate their utility in sustainable applications, including water treatment, renewable energy harvesting, and environmental remediation.

The Importance of Structured Water in HDAC8 for Correct Binding Pose Applied for Drug Design of Anticancer Molecules.

Validating the docking procedure and maintaining the structural water molecules at HDAC8 catalytic site. Molecular docking simulations play a significant role in Computer-Aided Drug Design, contributing to the development of new molecules. To ensure the reliability of these simulations, a validation process called "Re-docking" is employed, focusing on the binding mode of a ligand co-crystallized with the protein of interest. In this study, several molecular docking studies were conducted using five X-ray structures of HDAC8-ligand complexes from the PDB. Ligands initially complexed with HDAC8 were removed and re-docked onto the free protein, revealing a poor reproduction of the expected binding mode. In response to this, we observed that most HDAC8-ligand complexes contained one to two water molecules in the catalytic site, which were crucial for maintaining the cocrystallized ligand. These water molecules enhance the binding mode of the co-crystallized ligand by stabilizing the proteinligand complex through hydrogen bond interactions between ligand and water molecules. Notably, these interactions are lost if water molecules are removed, as is often done in classical docking methodologies. Considering this, molecular docking simulations were repeated, both with and without one or two conserved water molecules near Zn+2 in the catalytic cavity. Simulations indicated that replicating the native binding pose of co-crystallized ligands on free HDAC8 without these water molecules was challenging, showing greater coordinate displacements (RMSD) compared to those including conserved water molecules from crystals. The study highlighted the importance of conserved water molecules within the active site, as their presence significantly influenced the successful reproduction of the ligands' native binding modes. The results suggest an optimal molecular docking procedure for validating methods suitable for filtering new HDAC8 inhibitors for future experimental assays.

Analysis of multi-objective optimization applied to the simulation of a natural gas separation plant

This study concentrated on simulating and optimizing a natural gas separation plant using Unisim Design® software. The optimization procedure aimed to maximize gas production and the higher-molecular-weight hydrocarbons (yC5+) in the produced gas. The maximum water dew point of −45 °C of the produced gas was specified. The Nelder Mead/Weight Sum Method (NM/WSM) algorithm was employed in Python to optimize five decision variables involving the temperature and pressure of the main process streams. Results indicated that the relative deviations below 1%, compared to reference values, were obtained for the main process variables. The multi-objective optimization (MOO) results showed an increase in gas production of up to 5.58% (maximum value of 62,736.10 kg/h) and a rise in the molar fraction of yC5+ by up to 32.2% compared to the reference values. However, it was noted that maximizing gas production indirectly maximizes the yC5+, with only a single decision variable differing in the optimal direction, represented by the intermediate temperature of one compression stage. A Pareto curve was built to display the range of optimal solutions, showing variations of 0.035% in optimal gas production and 0.11% in yC5+. Furthermore, various simulation scenarios were optimized with different thermodynamic conditions at the unit's inlet. In all cases, the optimization resulted in an average gain of 1.57% in gas production regarding the reference (non-optimal case).

Simulation and optimization of EMI filter of conducted emission for high voltage gain DC-DC converter

In the continuity of the work on design of high voltage gain DC-DC converter, used for feeding the CubeSat electrospray thrusters with a variable output voltage according to the maneuver size, this paper introduces a simulation of the converter model with MATLAB/Simulink to determine its conducted emission level. By referring to the MIL-STD-461 standard, a required passive electromagnetic interference (EMI) filter parameters are calculated to reduce, under thresholds, the converter noise. With consideration of the available volume and power constraints in this kind of satellite, the design of common-mode (CM) choke is optimized with proposed procedure optimization, so as to reduce its volume and electrical losses. Also, this optimization procedure can be generalized and applied for any passive EMI filter design.

TSSG-CNN: A Tuberculosis Semantic Segmentation-Guided Model for Detecting and Diagnosis Using the Adaptive Convolutional Neural Network.

Tuberculosis (TB) is an infectious disease caused by Mycobacterium. It primarily impacts the lungs but can also endanger other organs, such as the renal system, spine, and brain. When an infected individual sneezes, coughs, or speaks, the virus can spread through the air, which contributes to its high contagiousness. The goal is to enhance detection recognition with an X-ray image dataset. This paper proposed a novel approach, named the Tuberculosis Segmentation-Guided Diagnosis Model (TSSG-CNN) for Detecting Tuberculosis, using a combined semantic segmentation and adaptive convolutional neural network (CNN) architecture. The proposed approach is distinguished from most of the previously proposed approaches in that it uses the combination of a deep learning segmentation model with a follow-up classification model based on CNN layers to segment chest X-ray images more precisely as well as to improve the diagnosis of TB. It contrasts with other approaches like ILCM, which is optimized for sequential learning, and explainable AI approaches, which focus on explanations. Moreover, our model is beneficial for the simplified procedure of feature optimization from the perspectives of approach using the Mayfly Algorithm (MA). Other models, including simple CNN, Batch Normalized CNN (BN-CNN), and Dense CNN (DCNN), are also evaluated on this dataset to evaluate the effectiveness of the proposed approach. The performance of the TSSG-CNN model outperformed all the models with an impressive accuracy of 98.75% and an F1 score of 98.70%. The evaluation findings demonstrate how well the deep learning segmentation model works and the potential for further research. The results suggest that this is the most accurate strategy and highlight the potential of the TSSG-CNN Model as a useful technique for precise and early diagnosis of TB.

From exergoeconomics to Thermo-X Optimization in the transition to sustainable energy systems

Exergoeconomics has played an important role in the study of new energy systems in the last decades. The use of exergy as a “carrier of value” has made it possible to define an unambiguous criterion for allocating costs among energy systems products. The paper shows how exergoecomic procedures of analytical optimization and design improvement on heuristic basis have been progressively replaced by more efficient procedures aimed at minimizing an objective cost function, leaving exergoecomic cost evaluation as the final step. It also highlights how growing concerns about climate change and the continued growth of inequalities in energy availability have broadened the set of objectives in the search for the optimal integration of the design and operation of energy conversion units and network with intelligent methodologies to reduce energy demands. The objective of the article is twofold: i) present the evolution of the main Exergoeconomic methods and show how they have paved the way for Thermo-X Optimization methods; and ii) outline the path for developing a general model of society's entire energy system that includes a broader set of objectives and constraints in addition to economic ones to help build the energy system of the future with a more sustainable perspective.

Optimal Rehabilitation Procedure for Intermittent Water Supply Systems

Optimal Rehabilitation Procedure for Intermittent Water Supply Systems

Optimal Uniform Strength Design of Frame and Lattice Structures

This paper provides a procedure to obtain the uniform strength of frame and lattice structures. Uniform strength condition is achieved by performing the shape optimization of all beam elements of the structure. The beam shape which guarantees uniform strength is analytically deduced from the one-dimensional Timoshenko model. The optimization problem presents itself as the search for the zeros of the objective-functions vector, which is a non-linear system of equations representing the kinematic-congruence and forces balance at every node of the structure. The analytical formulation of the optimization problem allows to construct the objective-functions vector without the use of external structural computation, i.e. not recurring to any Finite Element Analysis to accomplish iterations. This latter feature entails a great advantage in terms of computing time required to perform optimization. The proposed analytical formulation allows to directly insert the uniform strength condition into the objective-functions vector, transforming the optimization into an unconstrained problem. Some examples are shown in which the performance of the optimization procedure is discussed in terms of robustness and rate of computational complexity while increasing the degrees of freedom of the structure. The reliability and the quality of the optimization are verified through Finite Element Analysis.

Experimental and numerical study on moisture diffusion behavior of 3D woven composite

In this paper, a meso-scale finite element model was established to investigate the moisture diffusion in 3D woven carbon fiber reinforced composites (CFRCs), and to analyze the internal stresses caused by hygroscopic deformation. To quickly obtain the mesoscopic moisture diffusion parameters, an optimization procedure that fitted the simulation results to the experimental results was implemented to achieve automatic inverse identification of each parameter. The homogenized analytical models for describing the moisture diffusion in yarns and resin, specifically Fick’s and Langmuir-type models, were evaluated. The results indicated that Langmuir-type model provided more accurate fitting results, attributed to its consideration of anomalous moisture absorption. Furthermore, the internal stress induced by moisture absorption was analyzed for composites aging at 82 °C over different time intervals. The accumulation of hygroscopic stress was observed at the intersection of warp and weft yarns, as well as at the interface between the weft yarn and matrix.

Optimizing a machine learning design of dielectric properties in lead-free piezoelectric ceramics

Designing lead-free piezoelectric ceramics with tailored electrical properties remains a critical challenge for various applications. In this paper we present a novel methodology integrating Machine Learning (ML) and optimization procedures to fine-tune electrical properties in lead-free (1-x) Na0.5 Bi0.5 TiO3 - x CaTiO3 piezoelectric ceramics. A comprehensive dataset of dielectric measurements serves as the foundation for training ML models that accurately predict the permittivity (ε′) and dielectric loss (tan δ) as functions of Ca2+ concentration (x % Ca), temperature and frequency. Two ML techniques are evaluated: random forest regression, and Multi-Layer Perceptron neural network Regression (MLPR). The MLPR model exhibited a superior regression performance, achieving a correlation coefficient of 0.931 and a root mean squared error of 0.029. The MLPR was then optimized by the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to maximizes ε′ while minimizes tan δ. Within the NSGA-II framework, the optimal values were found at the Pareto curve knee, corresponding to a frequency, temperature, and x % Ca of 609.739 kHz, 398.15 K, and 6.10, respectively, resulting in ε′ equal to 857.87 and tan δ equal to 0.0120. This approach demonstrates the effectiveness of combining ML and optimization for designing the electrical properties of piezoelectric ceramics, paving the way for more efficient and targeted material development.