ADVANCED PREDICTIVE MODELING FOR ENHANCING MANUFACTURING EFFICIENCY IN CONCRETE STRUCTURES: A NOVEL HYBRID APPROACH

Abstract

Concrete structures are fundamental to modern infrastructure, and their efficient manufacturing is crucial for sustainable construction practices. However, traditional manufacturing processes often lack the precision and optimization required to meet evolving structural demands and sustainability goals. This deficiency becomes even more critical when considering seismic hazards, which pose a significant risk to the safety and resilience of urban infrastructure, particularly reinforced concrete buildings. Accurate assessment of seismic safety is crucial for effective risk mitigation and disaster preparedness. In this study, we introduce a novel approach that leverages a Fine-tuned Dragonfly Optimized Artificial Neural Network (FDO-ANN) to enhance the evaluation of seismic hazard safety in concrete Structures, utilizing data from the Structural Engineering Research Unit (SERU) database. Z-score normalization was employed as a data preprocessing approach to ensure the accuracy and reliability of the data utilized in the evaluation. Linear Discriminant Analysis (LDA) was used for feature extraction to identify essential characteristics or characteristics in reinforced concrete buildings that are associated with seismic safety. Python tool was used to analyze the proposed method. The proposed approach is assessed in terms of various parameters and compared to existing methods achieving an impressive accuracy of 95.6%. The proposed approach has the potential to inform more effective mitigation strategies, leading to increased resilience in the face of seismic hazards and improved protection of human lives and property.

Keywords

• manufacturing
• concrete structures
• seismic hazard
• safety
• fine-tuned dragonfly optimized artificial neural network (ftd-ann)