Wyniki wyszukiwania - BazTech

1

Analyzing the Impact of Population Size in AI-Based Reconstruction of the Thermal Parameter in Heat Conduction Modeling

Gawrońska Elżbieta, Zych Maria, Dyja Robert, Kowalkowski Michał

Advances in Science and Technology. Research Journal

|

2024

|

Vol. 18, no 2

349--364

EN

The research shows how to use swarming algorithms to rebuild the heat transfer coefficient, especially in regard to the continuous border condition. The authors utilized their application software to do numerical computations, employing classical variants of swarm algorithms. The numerical calculations employed a functional determining error to assess the accuracy of the estimated result. The functional minimization was conducted with the swarm algorithms (especially ABC and ACO). The geometry analyzed in this study consisted of a square shape referred to as the cast, enclosed within another square shape known as the casting mold. These two squares were separated by a layer facilitating heat conduction, characterized by the coefficient κ. The coefficient of the thermally conductive layer was recalibrated utilizing swarm methods within the range of 900 - 1500 [W/m^2K] and subsequently compared to a predetermined reference value. A finite element mesh consisting of 576 nodes was used for the calculations. The study involved simulations with populations of 5, 10, 15, and 20 individuals. Furthermore, each scenario also took into account noise of 0%, 2%, and 5% of the reference values. Results make evident the reconstructed value of the κ coefficient, cooling curves, and temperatures for the ABC and ACO algorithms are physically correct. The consequences indicate a notable level of satisfaction and strong concurrence with the anticipated of the κ parameter values. The results from the numerical simulations demonstrate considerable promise for applying artificial intelligence algorithms in optimizing production processes, analyzing data, and facilitating data-driven decision-making.

2

Impact of input parameters on numerical calculations optimized by swarming algorithms during computer simulations of the heat conduction phenomenon

Zych Maria, Gawrońska Elżbieta, Dyja Robert, Krawiec Piotr

Journal of Applied Mathematics and Computational Mechanics

|

2022

|

Vol. 21, nr 4

107--118

EN

The article presents the application of swarming algorithms in heat conduction, taking into account the continuity of the boundary condition (type IV). The influence of the input parameters of the bee and ant algorithm and tessellation on the selection of the heat conduction coefficient between the casting mold and the casting in computer simulations was presented. The results were compared for two different finite element grids, a different number of individuals, and a different number of iterations. The study also considered the magnitude of the reference temperature disturbance as the input temperature for numerical calculations. The analysis showed that the relative error of reproducing the value of the thermal conductivity coefficient in the continuity condition did not exceed 1.5% of the reference value of this coefficient.

3

Selection of the heat transfer coefficient using swarming algorithms

Gawrońska Elżbieta, Dyja Robert, Zych Maria, Domek Grzegorz

Acta Mechanica et Automatica

|

2022

|

Vol. 16, no. 4

325--339

EN

The article presents the use of swarming algorithms in selecting the heat transfer coefficient, taking into account the boundary condition of the IV types. Numerical calculations were made using the proprietary TalyFEM program and classic form of swarming algorithms. A function was also used for the calculations, which, during the calculation, determined the error of the approximate solution and was minimalised using a pair of individually employed algorithms, namely artificial bee colony (ABC) and ant colony optimisation (ACO). The tests were carried out to select the heat transfer coefficient from one range. Describing the geometry for a mesh of 408 fine elements with 214 nodes, the research carried out presents two squares (one on top of the other) separated by a heat transfer layer with a κ coefficient. A type III boundary condition was established on the right and left of both edges. The upper and lower edges were isolated, and a type IV boundary condition with imperfect contact was established between the squares. Calculations were made for ABC and ACO, respectively, for populations equal to 20, 40 and 60 individuals and 2, 6 and 12 iterations. In addition, in each case, 0%, 1%, 2% and 5% noise of the reference values were also considered. The obtained results are satisfactory and very close to the reference values of the κ parameter. The obtained results demonstrate the possibility of using artificial intelligence (AI) algorithms to reconstruct the IV type boundary condition value during heat conduction modelling.