Research on Modelling Architectural Heritage of Third-line Construction based on Hierarchical Analysis and Data Fusion using Rat Swarm tuned Artificial Neural Network

Abstract

This study explores the novel use of Artificial Neural Networks (ANNs) with Rat Swarm Optimization (ANN-RSO) to model the architectural legacy of third-line construction projects. A multi-layered, combined approach to thoroughly assess and protect valuable historical buildings is constructed by the study through the application of hierarchical analysis and data fusion tools. The approach makes use of ANN-RSO power to maximize the analysis and understanding of a variety of data sets, from historical and cultural relevance to structural details and material compositions. Through the systematic division of complicated data into digestible layers, hierarchical analysis improves the neural network's power by concentrating on distinct aspects at various levels. Data fusion combines different data kinds at the same time, such as verbal descriptions, architectural plans, and photographic proof, to create a rich, consistent database that feeds into the neural network. By fine-tuning the ANN parameters, the RSO approach greatly increases the model's efficiency and accuracy in predicting and modeling architectural features. This study provides a standard for the use of modern computational techniques in the protection of cultural heritage in addition to showcasing the potential of ANN-RSO in architectural heritage modeling. The results show that these technologically advanced models can be essential resources for architectural historians and preservationists working on reconstruction projects, giving them a better understanding of the past and more precise reconstructions of historical buildings.