Optimization of Steel Frame Structures using Genetic Algorithm
Аннотация
Compact structural components are required in civil engineering, particularly for uses like building modules and the restoration of damaged flooring in current structures. Beamcolumn links significantly influence steel frame architecture and assessment outcomes. Due to the need to equilibrium the mass and cost of these constructions, designing a hybrid steel frame construction is frequently difficult. Perforated steel sandwiched components must have exceptional compositional efficiency in terms of impact resilience when used on ships. This is frequently accomplished by adding weight without fully utilizing the structure's properties. A MATLAB-developed multifaceted optimum technique using a back-propagation (BP) neural network (NN) and a genetic algorithm (GA) is shown here, taking into account the physical characteristics of sandwiched panels beneath static and impacts. Energy intake, architectural weight, static and flexible strain, and distortion were the assessment factors for this approach. Using quantitative model computations, suitable sample values were acquired before optimization. The BP NN was subsequently utilized to produce an operational connection between the output factors and the layout information. In order to get the optimal outcome, a multifaceted optimization study was conducted using the specified functionality and both a conventional and an adaptable GA. Employing a small number of samples, this research produced an innovative idea that is highly reliable and efficient for identifying the architectural characteristics that offer the greatest durability.