Hybrid Krill Herd-ANN Model for Prediction Strength and Stiffness of Bolted Connections

The behavior of beam-to-column connections significantly influences the stability, strength, and stiffness steel structures. This is particularly important in extreme non-elastic responses, i.e., earthquakes, sudden column removal, as fluctuation strength affects both supply demand. Accordingly, it essential to accurately estimate analysis design procedures for Beginning with state-of-the-art, capacity three available component-based mechanical models complex properties top- seat-angle double-web angles (TSACWs), variable parameters, were investigated. Subsequently, a novel hybrid krill herd algorithm-artificial neural network (KHA-ANN) model was proposed acquire an informational from experimental dataset. Using several statistical metrics, including corresponding coefficient variation (CoV), correlation (R), provided by Taylor diagram, this study revealed that herd-ANN achieved most reliable predictive accuracy double web angles.