Linked mathematical–informational modeling of perforated steel plate shear walls

• Represent an approach to the prediction of force-displacement curves by mathematical modeling.
• The learning capabilities of neural networks are used to model the pinched behavior of perforated shear walls.
• A neural network structure is developed to model the path dependency of the shear wall behavior.
• By employing behavioral models to estimate the F–D behavior, analytical expressions for evaluating major shear wall parameters, such as initial stiffness and ultimate force, are derived.

Steel plate shear walls have come to be considered as an appropriate system for resisting lateral loads due to earthquakes and wind, especially in tall structures, for their flexible, energy dissipation and suitable post-buckling behaviors. This paper presents prediction ways for the mechanical behavior of shear walls. In this study, two different methods are suggested to consider the intricate hysteretic behavior of perforated shear walls. To use the benefits of both mathematical and informational presentations, a novel method, a linked modeling skeleton, is created and demonstrated through representing the complex behavior of perforated shear walls.