Integrating Artificial Intelligence in Earthquake-Resistant Structural Design - Advances in Sustainable Building and Concrete Technology

Smart building management and construction is tremendous for developing smart cities in building sites and is known for its stability and durability. However, its performance can be significantly enhanced by improving material properties such as strength, fire resistance and impact protection. Conventional earthquake structural design considers only a limited number of factors, mainly elastic structural properties, to determine the critical design parameters.Yet, these parameters are often suboptimal since they do not consider the extensive plasticity expected in building structures during earthquakes. One significant challenge in concrete design is that it is difficult to predict the exact performance of a particular concrete mix without extensive testing, which is time-consuming and costly.Conventional techniques for optimizing concrete properties depend significantly on empirical testing and expert intuition, which are time-consuming and may not completely handle the complex interactions among various material components.To address the above problems, this research presents the Artificial Intelligence (AI) based Multi-Layer Perceptron Neural Network (MLPNN) method for efficient building construction that resists earthquakes.To start with the proposed work, C-Score Normalization (CSN) method is employed to normalize the collective dataset. Then, select essential features of concrete materials using the Deep Feature Elimination with Residual Network (DFE-RN)approach. Following that, the MLPNN method is used to classify the best materials for efficient building construction that resists earthquakes. The proposed framework has the potential to revolutionize the building industry by constructing concrete with improved properties, reducing the need for extensive physical testing and speeding up the innovation process.This paper demonstrates the proposed AI-based approach can effectively improve Earthquake-resistant structural design. The proposed simulation result illustrates the efficient performance regarding precision, recall, classification accuracy and F1-score with less time complexity.