Modelling Dimensional Accuracy and Surface Roughness in Resin Additive Manufacturing through Neural Network: A Multi-objective Optimization Approach in Dentistry

Journal of Materials Engineering and Performance, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 11, 2025

Language: Английский

---

Explainable AI Techniques for Comprehensive Analysis of the Relationship between Process Parameters and Material Properties in FDM-Based 3D-Printed Biocomposites

Journal of Manufacturing and Materials Processing, Journal Year: 2024, Volume and Issue: 8(4), P. 171 - 171

Published: Aug. 6, 2024

This study investigates the complex relationships between process parameters and material properties in FDM-based 3D-printed biocomposites using explainable AI techniques. We examine effects of key parameters, including biochar content (BC), layer thickness (LT), raster angle (RA), infill pattern (IP), density (ID), on tensile, flexural, impact strengths FDM-printed pure PLA biochar-reinforced composites. Mechanical testing was used to measure ultimate tensile strength (UTS), flexural (FS), (IS) samples. The extreme gradient boosting (XGB) algorithm build a predictive model based data collected from mechanical testing. Shapley Additive Explanations (SHAP), Local Interpretable Model-Agnostic (LIME), Partial Dependence Plot (PDP) techniques were implemented understand interactions such as UTS, FS, IS. Prediction by XGB accurate for IS, with R-squared values 0.96, 0.95, 0.85, respectively. explanation showed that has most significant influence UTS SHAP +2.75 +5.8, BC value +2.69. PDP reveals 0.3 mm LT 30° RA enhances properties. contributes field application artificial intelligence additive manufacturing. A novel approach is presented which machine learning XAI SHAP, LIME, are combined not only optimization but also provide valuable insights about interaction

Language: Английский

---

Experimental Study and Random Forest Machine Learning of Surface Roughness for a Typical Laser Powder Bed Fusion Al Alloy

Metals, Journal Year: 2024, Volume and Issue: 14(10), P. 1148 - 1148

Published: Oct. 8, 2024

Surface quality represents a critical challenge in additive manufacturing (AM), with surface roughness serving as key parameter that influences this aspect. In the aerospace industry, of aviation components is very important parameter. study, typical Al alloy, AlSi10Mg, was selected to study its when using Laser Powder Bed Fusion (LPBF). Two Random Forest (RF) models were established predict upper printed samples based on laser power, scanning speed, and hatch distance. Through it found two-dimensional (2D) RF model successful predicting values experimental data. The best minimum 2.98 μm, which known without remelting. More than two-thirds had less 7.7 μm. maximum 11.28 And coefficient determination (R2) 0.9, also suggesting 3D-printed alloys can be predicted ML approaches such model. This helps understand relationship between printing parameters print better quality.

Language: Английский

---

Neuro-Fuzzy Model Evaluation for Enhanced Prediction of Mechanical Properties in AM Specimens

Applied Sciences, Journal Year: 2024, Volume and Issue: 15(1), P. 7 - 7

Published: Dec. 24, 2024

This paper explores the integration of adaptive neuro-fuzzy inference systems (ANFIS) with additive manufacturing (AM) to enhance prediction mechanical properties in 3D-printed components. Despite AM’s versatility producing complex geometries, achieving consistent performance remains challenging due various process parameters and anisotropic behavior printed parts. The proposed approach combines learning capabilities neural networks decision-making strengths fuzzy logic, enabling ANFIS refine printing improve part quality. Experimental data collected from AM processes are used train model, allowing it predict outputs such as stress, strain, Young’s modulus under values. predictive model was assessed root mean square error (RMSE) coefficient determination (R2) evaluation metrics. study initially examined impact key on subsequently compared two partitioning techniques—grid subtractive clustering—to identify most effective configuration. experimental results analysis demonstrated that could dynamically adjust parameters, leading significant improvements accuracy modulus, showcasing its potential address inherent complexities processes.

Language: Английский

---

A Fuzzy Inference System for Predicting Air Traffic Demand based on Socioeconomic Drivers

Saudi Journal of Engineering and Technology, Journal Year: 2024, Volume and Issue: 9(08), P. 377 - 388

Published: Aug. 14, 2024

The past ten years have seen significant expansion in the aviation sector, which during previous five has steadily pushed emerging countries closer to economic independence. It is crucial accurately forecast potential demand for air travel make long-term financial plans. To market low-cost passenger carriers, this study suggests working with airlines, airports, consultancies, and governmental institutions' strategic planning divisions. aims develop an artificial intelligence-based methods, notably fuzzy inference systems (FIS), determine most accurate forecasting technique domestic carrier Bangladesh. give end users real-world applications, includes nine variables, two sub-FIS, one final Mamdani Fuzzy Inference System utilizing a Graphical User Interface (GUI) made app designer tool. evaluation criteria used inquiry included mean square error (MSE), accuracy, precision, sensitivity, specificity. effectiveness of developed Air Passenger Demand Prediction FIS assessed using 240 data sets, specificity, MSE values are 90.83%, 91.09%, 90.77%, 2.09%, respectively.

Language: Английский

---