Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
| Autor: | Diego González-Aguilera, Roberto Garcia-Martin, José G. Fueyo, Francisco J. Madruga, Javier Pisonero, Manuel Rodríguez-Martín, Ángel Luis Muñoz |
|---|---|
| Přispěvatelé: | Universidad de Cantabria |
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
active thermography (AT)
quality assessment (QA) Computer science Gaussian 02 engineering and technology Machine learning computer.software_genre additive materials (AM) lcsh:Chemical technology 01 natural sciences Biochemistry Article Analytical Chemistry 010309 optics symbols.namesake Additive materials (AM) Kriging Non-destructive testing (NDT) 0103 physical sciences Linear regression Feature (machine learning) lcsh:TP1-1185 Electrical and Electronic Engineering Instrumentation Finite element method (FEM) business.industry finite element method (FEM) machine learning (ML) Machine learning (ML) 021001 nanoscience & nanotechnology Atomic and Molecular Physics and Optics Random forest Support vector machine Active thermography (AT) Multilayer perceptron Thermography Outlier Quality assessment (QA) symbols non-destructive testing (NDT) Artificial intelligence 0210 nano-technology business computer |
| Zdroj: | Sensors, Vol 20, Iss 3982, p 3982 (2020) Sensors (Basel, Switzerland) Sensors, 2020, 20(14), 3982 UCrea Repositorio Abierto de la Universidad de Cantabria Universidad de Cantabria (UC) Sensors Volume 20 Issue 14 |
| ISSN: | 1424-8220 |
| Popis: | The present article addresses a generation of predictive models that assesses the thickness and length of internal defects in additive manufacturing materials. These modes use data from the application of active transient thermography numerical simulation. In this manner, the raised procedure is an ad-hoc hybrid method that integrates finite element simulation and machine learning models using different predictive feature sets and characteristics (i.e., regression, Gaussian regression, support vector machines, multilayer perceptron, and random forest). The performance results for each model were statistically analyzed, evaluated, and compared in terms of predictive performance, processing time, and outlier sensibility to facilitate the choice of a predictive method to obtain the thickness and length of an internal defect from thermographic monitoring. The best model to predictdefect thickness with six thermal features was interaction linear regression. To make predictive models for defect length and thickness, the best model was Gaussian process regression. However, models such as support vector machines also had significative advantages in terms of processing time and adequate performance for certain feature sets. In this way, the results showed that the predictive capability of some types of algorithms could allow for the detection and measurement of internal defects in materials produced by additive manufacturing using active thermography as a non-destructive test. This research was funded by Ministry of Science and Innovation, Government of Spain, through the research project titled Fusion of non-destructive technologies and numerical simulation methods for the inspection and monitoring of joints in new materials and additive manufacturing processes (FaTIMA) with code RTI2018-099850-B-I00. The authors are grateful to the Fundación Universidad de Salamanca for the indirect support provided by the ITACA proof-of-concept project (PC_TCUE_18-20_047), being this helpful for some of the purposes of this article. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|