ANN model for predicting the significance of parameters in the structural behaviour of an elliptical paraboloid shell roof

Abstract

In the analysis of various engineering problems, Finite Element Method (FEM) is considered as one of the significant tools and has been broadly used in many fields. Employing this method aids in simulating the deformable complex objects comprising material and structural properties and determines the model’s response to definite loading conditions. For an easier, quicker and less expensive analysis of structures, a consistent model has to be developed. However, many techniques optimize the system performance; the intelligent method design is a valuable technique in optimizing the competence of such structures. One of the emerging brilliant techniques is Artificial Neural Network (ANN) which is applied in modelling, simulation and system organization. To resolve difficult and nonlinear problems ANN tool is faster and more precise as compared to other conventional techniques. The present work is carried out to predict the behaviour of Elliptical Paraboloid Shell (EPS) roof based on the output obtained from an ANN model. The results indicate that the ANN technique performs faster and predicts the output with less error and is an appropriate method to model the shell roof structure’s performance. The influence of the geometrical parameters of the shell in the prediction of deflection has been studied.