Informational and hybrid-informational modelling of yielding shear panel device

Yielding Shear Panel Device (YSPD) is a recently developed passive control device, which is designed to exploit the shear deformation characteristics of a steel diaphragm plate to dissipate energy when subjected to seismic excitation. Force-displacement response for a wide variety of YSPDs obtained using validated Finite Element (FE) models are used in the current paper to formulate a mathematical model using Neural Network (NN) to facilitate the structural modelling of the device. The proposed NN informational model is able to accurately represent the complex pinching hysteretic response of YSPDs. The concept is further extended to devise a hybrid informational model to utilise the learning ability of a NN model within a mathematical framework. The proposed hybrid model shows good agreement in predicting the force-deformation response of YSPDs in a simplistic manner and should facilitate the full scale modelling of building frames retrofitted with YSPD without the need for complicated and detailed FE modelling of the device.