A Review on Artificial Neural Networks for Structural Analysis

PurposeArtificial neural networks are recently developed information processing-based methods that emerged as unique tools to analyse the behaviour of structures. The present study reviews work on the bending, buckling, and vibrations of beams, plates, shells, and panels using artificial neural networks. A detailed description of artificial neural networks, their learning process, and different methodologies has been provided here.MethodologyA discussion over multilayer perceptron algorithms is presented to optimize the efficiency and effectiveness of learning. The data is exported from Scopus for a well-defined period, and its rigorous examination is made to study the publication trends, citation patterns, geographical distribution, and primary focus and identify emerging interests which further provide the details of the development of ANNs, their limitations, and potential areas for future exploration in the field of structural engineering.ResultsThis paper provides a comprehensive literature review of the development and application of artificial neural networks to investigate the structural behaviour of beams, plates, and shells. Moreover, it also reports the interdisciplinary research areas along with advanced machine learning algorithms, big data analysis, computational techniques, and the exploration of new applications in emerging fields. Also, it discussed the developments, future scopes, advantages, disadvantages, challenges, and limitations in this field of study.