Machine learning and neural networks based approach for deflection prediction of Euler-Bernoulli beam equations

Abstract

Beam-like structures are widespread but essential systems that have been extensively studied for centuries. Although several proposed solutions are effective, the time consumption and the difficulty of reconstructing the problem are the major disadvantages of these methods. This paper offers a new methodology for finding solutions to beam problems based on Machine Learning and Neural Networks with different optimization algorithms. Various regression models are compared on numerically stimulated Euler-Bernoulli beam modelling.