George Markou, Nikolaos Bakas, Manolis Papadrakakis, Ashley van der Westhuizen, Duan Calitz, Victor T. Chibaya, Kevin T Braun

In our recent works, we have made significant strides in advancing the predictive modeling of structural behaviors through the use of artificial intelligence and machine learning techniques. In one study, we developed an Artificial Neural Network model capable of predicting the fundamental period of steel structures with an impressive accuracy of 99.99% correlation and a mean absolute percentage error of 0.7%, taking into account the crucial soil-structure interaction effects, particularly on soft soils. Another work in our portfolio includes the creation of predictive formulae for calculating the fundamental period of unbraced steel framed structures. By using machine learning, automated algorithms, and considering various geometric configurations and steel column members, we achieved a coefficient of determination value of 99.976% and an error less than 2%, even while incorporating the soil-structure interaction effects. Extending our research to the analysis of horizontally curved steel I-beams, we tackled the challenging task of predicting sectional rotations due to bending. Through the combination of 3D finite element analysis and higher-order nonlinear regression machine learning algorithms, we developed a novel predictive formula, achieving a mean absolute error of just 1.44%. Together, these studies underscore our commitment to enhancing the precision of structural behavior predictions, contributing to safer and more reliable design practices in the civil engineering domain. This project would be unfeasible to run on a standard personal computer, as it utilised 400,000 cpu hours in developing the required dataset that was used to derive the predictive models. Special thanks to MeluXina (https://docs.lxp.lu/) as well as Cyclone (https://hpcf.cyi.ac.cy/) supercomputers for providing access to the CPU partition.

• Markou, N. Bakas, S. Chatzichristofis, and M. Papadrakakis, “A general framework of high-performance machine learning algorithms: application in structural mechanics,” Comput. Mech. Accept. August 2023, doi: https://doi.org/10.1007/s00466-023-02386-9
• Duan Calitz, George Markou, Nikolaos Bakas, and Manolis Papadrakakis. “Developing fundamental period formulae for steel framed structures”. In: 9th International Conference on Computational Methods in Structural Dynamics and Earthquake Engineering. COMPDYN 2023, Athens, Greece, June 2023. url: https://2023.compdyn.org/.
• Vistor Chibaya, George Markou, Nikolaos Bakas, and Manolis Papadrakakis. “Development of formulae for the section rotations due to bending of curved steel i-beams through ai and ml algorithms”. In: 9th International Conference on Computational Methods in Structural Dynamics and Earthquake Engineering. COMPDYN 2023, Athens, Greece, June 2023. url: https://2023.compdyn.org/.
• Ashley Megan van der Westhuizen, George Markou, Nikolaos Bakas, and Manolis Papadrakakis. “Developing an artificial neural network model that predicts the fundamental period of steel structures using a large dataset”. In: 9th International Conference on Computational Methods in Structural Dynamics and Earthquake Engineering. COMPDYN 2023, Athens, Greece, June 2023. url: https: //2023.compdyn.org/.
• K T Braun, N Bakas, G Markou, and S W Jacobsz. “Advanced numerical modelling of the nonlinear mechanical behaviour of a laterally loaded pile embedded in stiff unsaturated clay”. en. In: Journal of the South African Institution of Civil Engineering 65 (June 2023), pp. 28–38. issn: 1021-2019. url: http://www.scielo.org.za/scielo.php?script=sci arttext&pid=S1021-20192023000200004&nrm=iso.
• Westhuizen, G. Markou, and N. Bakas, “Use of ai and ml algorithms in developing closed-form formulae, for structural engineering design,” in Artificial Intelligence and Machine Learning Techniques for Civil Engineering, 2023. [Online]. Available: https://www.igi-global.com/chapter/use-of-ai-and-ml-algorithms-in-developing-closed-form-formulae-for-structural-engineering-design/324541