VISCOELASTIC DOUBLE BEAM SYSTEM STABILITY ANALYSIS USING ARTIFICIAL NEURAL NETWORKS

This paper presents neural networks application and their advantages in estimation and analysis of numerical results obtained from previous research of dynamic stability of double beam system under stochastic loading processes. For determining the bounds of the almost sure stability of the double-beam system, direct Lyapunov method was applied and conventional numerical methods for stability region determination were used. Numerical results obtained by dynamic stability analysis were then used as a training data for artificial neural network (ANN). Application of ANN allows reduced computation time for accurate determination of stability regions, while the main advantage of trained ANN is that it estimates results even in cases where conventional numerical method fails