# The Study of Machine Learning Assisted the Design of Selected Composites Properties

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Work Methodology

- -
- by measuring, we will determine the absorption values of all material samples (VO_20_PVB_80_TF, VO_30_PVB_70_TF and VO_50_PVB_50_TF);
- -
- materials VO_20_PVB_80_TF and VO_50_PVB_50_TF will be tested in a climate chamber;
- -
- using the measured values, we will compile a training set of data in order to design models for the prediction of values for the material VO_30_PVB_70_TF, which is not tested in a climate chamber;
- -
- we will verify the results obtained through models for materials VO_20_PVB_80_TF and VO_50_PVB_50_TF separately by comparing them with their measured values;
- -
- the more favorable of the models will be applied to predict data for the material VO_30_PVB_70_TF.

#### 2.1. Materials and Test Characterization

^{−2}·nm

^{−1}and relative humidity. The whole cycle lasted 10 h. In this case, the samples were tested first for seven cycles 70 h and the total exposure time in the climate chamber was 32 cycles, i.e., 320 h. Absorption values (YLabel) for materials VO_20_PVB_80_TF, VO_30_PVB_70_TF, and VO_50_PVB_50_TF were measured using an equipment 620-IR Varian [15]. After the action of the climate chamber on the composite material, it can be seen that the values for certain bundles are different, i.e., that the material’s structure changes during the action of water, ultraviolet radiation, and condensation. The samples were placed in the instrument with a phase textile part to penetrate the transmission transmitted by a sensor mounted on a diamond tip. Subsequently, for the materials VO_20_PVB_80_TF and VO_50_PVB_50_TF, the absorption values after the stay in the climatic chamber were obtained for 70 h (designated KS1) and after 320 h (designated KS2). The number of records for each material was the value 1869 (Figure 3 and Figure 4).

#### 2.2. Methods and Tools Characterization

- Matlab Software Application Tools

- Machine Learning

- Regression Analysis

- Qualitative Evaluation

^{2}) is commonly used to Evaluate (1) the goodness of the linear fit of regression models in ANNs. A value of 1 means that the regression model explains all the predicted variables, which means that the correlation between the two variables is perfect [25].

## 3. Model Proposal

#### Neural Networks

^{2}and MSE are displayed from the qualitative characteristics. If the values do not meet our expectations, we have the opportunity to overtrain the data several times or change other factors affecting these values. After repeated training, we reached the deal R

^{2}= 0.922599. The following scheme shows (Figure 7) the possibility of a graphical representation of the selected model.

^{2}is 0.73 (Figure 8).

## 4. Results and Discussion

#### Application of Selected Model for VO_30_PVB_70_TF

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## References

- Chen, C.T.; Gu, G.X. Machine learning for composite materials. MRS Commun.
**2019**, 9, 556. [Google Scholar] [CrossRef] [Green Version] - Breque, M.; de Nul, L.; Petridis, A. Industry 5.0, Towards a Sustainable, Human-Centric and Resilient European Industry; European Commission: Brussels, Belgium, 2021. [Google Scholar]
- Industry 5.0. Available online: https://research-and-innovation.ec.europa.eu/research-area/industry/industry-50_en (accessed on 15 September 2022).
- Rodríguez-Martín, M.; Fueyo, J.G.; Gonzalez-Aguilera, D.; Madruga, F.J.; García-Martín, R.; Muñóz, Á.L.; Pisonero, J. Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods. Sensors
**2020**, 20, 3982. [Google Scholar] [CrossRef] - Furtado, C.; Pereira, L.F.; Tavares, R.P.; Salgado, M.; Otero, F.; Catalanotti, G.; Arteiro, A.; Bessa, M.A.; Camanho, P.P. A methodology to generate design allowables of composite laminates using machine learning. Int. J. Solids Struct.
**2021**, 233, 111095. [Google Scholar] [CrossRef] - Sharma, A.; Mukhopadhyay, T.; Rangappa, S.M.; Siengchin, S.; Kushvaha, V. Advances in Computational Intelligence of Polymer Composite Materials: Machine Learning Assisted Modeling, Analysis and Design. Arch. Comput. Methods Eng.
**2022**, 29, 3341–3385. [Google Scholar] [CrossRef] - Hrehova, S. Quality Evaluation of Heating Process Control Using Matlab Tools. In Proceedings of the 19th International Carpathian Control Conference (ICCC), IEEE, Szilvasvarad, Hungary, 28–31 May 2018. [Google Scholar] [CrossRef]
- Mathworks. Machine Learning. Available online: https://www.mathworks.com/help/stats/machine-learning-in-matlab.html (accessed on 13 March 2022).
- Artificial Neural Networks Overview. Available online: https://www.dataversity.net/artificial-neural-networks-overview/# (accessed on 11 March 2022).
- Paluszek, M.; Thomas, S. Practical MATLAB Deep Learning; Apress: Berkeley, CA, USA, 2020; p. 270. [Google Scholar]
- Hošovský, A.; Piteľ, J.; Trojanová, M.; Židek, K. Computational Intelligence in the Context of Industry 4.0. Implementing Industry 4.0 in SMEs; Palgrave Macmillan: Cham, Switzerland, 2021. [Google Scholar] [CrossRef]
- Bhattacharya, S.; Kalita, K.; Čep, R.; Chakraborty, S. A Comparative Analysis on Prediction Performance of Regression Models during Machining of Composite Materials. Materials
**2021**, 14, 6689. [Google Scholar] [CrossRef] [PubMed] - Sinay, J.; Kotianova, Z.; Balazikova, M.; Dulebová, M.; Markulik, Š. Measurement of low-frequency noise during CNC machining and its assessment. Measurement
**2018**, 119, 190–195. [Google Scholar] [CrossRef] - Emmert-Streib, F.; Dehmer, M. Evaluation of Regression Models: Model Assessment, Model Selection and Generalization Error. Mach. Learn. Knowl. Extr.
**2019**, 1, 521–551. [Google Scholar] [CrossRef] [Green Version] - Inyurt, S.; Kashani, M.H.; Sekertekin, A. Ionospheric TEC forecasting using Gaussian Process Regression (GPR) and Multiple Linear Regression (MLR) in Turkey. Astrophys. Space Sci.
**2020**, 365, 99. [Google Scholar] [CrossRef] - Nagyova, A.; Pacaiova, H.; Markulik, S.; Turisová, R.; Kozel, R.; Džugan, J. Design of a Model for Risk Reduction in Project Management in Small and Medium-Sized Enterprises. Symetry
**2021**, 13, 763. [Google Scholar] [CrossRef] - Priya, K.S. Linear Regression Algorithm in Machine Learning through MATLAB. Int. J. Res. Appl. Sci. Eng. Technol. (IJRASET)
**2021**, 9, 12. [Google Scholar] [CrossRef] - Hamidi, Y.K.; Berrado, A.; Altan, M.C. Machine learning applications in polymer composites. AIP Conf. Proc.
**2020**, 2205, 20031. [Google Scholar] [CrossRef] - Anuar, M.A.R.B.K.; Ngamkhanong, C.; Wu, Y.; Kaewunruen, S. Recycled Aggregates Concrete Compressive Strength Prediction Using Artificial Neural Networks (ANNs). Infrastructures
**2021**, 6, 17. [Google Scholar] [CrossRef] - Cvitic’, I.; Perakovic’, D.; Perisa, M.; Gupta, B.B. Ensemble machine learning approach for classification of IoT devices in smart home. Int. J. Mach. Learn. Cybern. Ensemble
**2021**, 12, 3179–3202. [Google Scholar] [CrossRef] - Jierula, A.; Wang, S.; Oh, T.-M.; Wang, P. Study on Accuracy Metrics for Evaluating the Predictions of Damage Locations in Deep Piles Using Artificial Neural Networks with Acoustic Emission Data. Appl. Sci.
**2021**, 11, 2314. [Google Scholar] [CrossRef] - Straka, M.; Rosová, A.; Radim, L.; Petr, B.; Janka, Š. Principles of computer simulation design for the needs of improvement of the raw materials combined transport system. Acta Montan. Slovaca
**2018**, 23, 163–174. [Google Scholar] - Malindzakova, M.; Straka, M.; Rosova, A.; Kanuchova, M.; Trebuna, P. Modeling the process for incineration of municipal waste. Przem. Chem.
**2015**, 94, 1260–1264. [Google Scholar] [CrossRef] - Wicher, P.; Staš, D.; Karkula, M.; Lenort, R.; Besta, P. A computer simulation-based analysis of supply chains resilience in industrial environment. Metalurgija
**2015**, 54, 703–706. [Google Scholar] - Szavai, S.; Kovacs, S.; Bezi, Z.; Kozak, D. Coupled Numerical Method for Rolling Contact Fatigue Analysis. Teh. Vjesn. Tech. Gaz.
**2021**, 28, 1560–1567. [Google Scholar] [CrossRef] - Gupta, B.B.; Tewari, A.; Cvitic, I.; Perakovic, D.; Chang, X.J. Articial intelligence empowered emals classifier for Internet of Things based system in industry 4.0. Wirel. Netw.
**2022**, 28, 493–503. [Google Scholar] [CrossRef] - Cvitic´, I.; Perakovic´, D.; Periša, M.; Botica, M. Novel approach for detection of IoT generated DDoS traffic. Wirel. Netw.
**2019**, 27, 1573–1586. [Google Scholar] [CrossRef] - Gubbi, J.; Buyya, R.; Marusic, S.; Palaniswami, M. Internet of Things (IoT): A vision, architectural elements, and future directions. Future Gener. Comput. Syst.
**2013**, 29, 1645–1660. [Google Scholar] [CrossRef] [Green Version] - Atzori, L.; Iera, A.; Morabito, G. The Internet of Things: A survey. Comput. Netw.
**2010**, 54, 2787–2805. [Google Scholar] [CrossRef] - Georgala, K.; Kosmopoulos, A.; Paliouras, G. Spam filtering: An active learning approach using incremental clustering. In Proceedings of the 4th International Conference on Web Intelligence, Mining and Semantics, Thessaloniki, Greece, 2–4 June 2014. [Google Scholar]
- Xie, Z.; Peng, J.; Sorokina, M.; Zeng, H. Design of Mode-Locked Fibre Laser with Non-Linear Power and Spectrum Width Transfer Functions with a Power Threshold. Appl. Sci.
**2022**, 12, 10318. [Google Scholar] [CrossRef] - Behunova, A.; Soltysova, Z.; Behun, M. Complexity Management and its impact on economy. TEM J. Technol. Educ. Manag. Inform.
**2018**, 7, 324–329. [Google Scholar] [CrossRef] - Grabara, J.K.; Dima, I.C.; Kot, S.; Kwiatkowska, J. Case on in-house logistics modeling and simulation. Res. J. Appl. Sci.
**2011**, 6, 7–12. [Google Scholar] [CrossRef] [Green Version] - Gor, M.; Dobriyal, A.; Wankhede, V.; Sahlot, P.; Grzelak, K.; Kluczyński, J.; Łuszczek, J. Density Prediction in Powder Bed Fusion Additive Manufacturing: Machine Learning-Based Techniques. Appl. Sci.
**2022**, 12, 7271. [Google Scholar] [CrossRef] - Kot, S.; Slusarczyk, B. Process simulation in supply chain using logware software. Ann. Univ. Apulensis Ser. Oeconomica
**2009**, 11, 932. [Google Scholar] - Schmid, J.; Trabesinger, S.; Brillinger, M.; Pichler, R.; Wurzinger, J.; Ciumasu, R. Tacit Knowledge Based Acquisition of Verified Machining Data. In Proceedings of the IEEE, 9th International Conference on Industrial Technology and Management (ICITM 2020), Oxford, UK, 11–13 February 2020; pp. 117–121. [Google Scholar]
- Yu, J.; Oh, S.J.; Baek, S.; Kim, J.; Lee, T. Predicting the Effect of Processing Parameters on Caliber-Rolled Mg Alloys through Machine Learning. Appl. Sci.
**2022**, 12, 10646. [Google Scholar] [CrossRef]

**Figure 1.**Three pillars of Industry 5.0 [2].

**Figure 5.**Machine learning dividing [10].

Method | Neural Net Fitting | Regression Learner | ||||||
---|---|---|---|---|---|---|---|---|

Material | VO-20-PVB-80-TF | VO-50-PVB-50-TF | VO-20-PVB-80-TF | VO-50-PVB-50-TF | ||||

State | KS1 | KS2 | KS1 | KS2 | KS1 | KS2 | KS1 | KS2 |

R^{2} | 0.9052 | 0.9866 | 0.7302 | 0.8655 | 0.7617 | 0.6325 | 0.6448 | 0.5659 |

MAPE | 8.56% | 11.21% | 12.04% | 15.37% | 14.92% | 27.11% | 18.64% | 29.45% |

MSE | 0.00094 | 0.000708 | 0.00359 | 0.00209 | 0.00237 | 0.00432 | 0.00473 | 0.00188 |

Values | ||||||||
---|---|---|---|---|---|---|---|---|

Wave | 399.19 | 401.12 | 403.05 | 404.97 | 406.90 | 408.83 | 410.76 | 412.69 |

Absorbance | 0.774 | 0.778 | 0.777 | 0.769 | 0.762 | 0.757 | 0.762 | 0.772 |

Values | ||||||||
---|---|---|---|---|---|---|---|---|

Absorbance | 0.774 | 0.778 | 0.777 | 0.769 | 0.762 | 0.757 | 0.762 | 0.772 |

KS1 | 0.5702 | 0.5707 | 0.5705 | 0.5692 | 0.5676 | 0.5659 | 0.5675 | 0.5697 |

KS2 | 0.6809 | 0.6865 | 0.6850 | 0.6702 | 0.6495 | 0.6288 | 0.6482 | 0.6759 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Hrehova, S.; Knapcikova, L.
The Study of Machine Learning Assisted the Design of Selected Composites Properties. *Appl. Sci.* **2022**, *12*, 10863.
https://doi.org/10.3390/app122110863

**AMA Style**

Hrehova S, Knapcikova L.
The Study of Machine Learning Assisted the Design of Selected Composites Properties. *Applied Sciences*. 2022; 12(21):10863.
https://doi.org/10.3390/app122110863

**Chicago/Turabian Style**

Hrehova, Stella, and Lucia Knapcikova.
2022. "The Study of Machine Learning Assisted the Design of Selected Composites Properties" *Applied Sciences* 12, no. 21: 10863.
https://doi.org/10.3390/app122110863