Smart design of overhead cranes using artificial intelligence and multi-objective optimization via evolutionary algorithms

Traditional overhead crane designs are known to be conservative, resulting in increased crane weight, wasteful material utilization, higher energy consumption, and higher operational expenses, which ultimately contribute to resource depletion and pollutant emissions. This study explores the optimization of overhead crane structures to minimize weight while fulfilling essential performance requirements. A comprehensive optimization framework of overhead crane design to achieve the least weight while satisfying performance constraints. A fractional central composite design was employed to generate various designs, which were then simulated using finite element analysis (FEA) in Ansys Workbench. The resulting data were utilized to train an artificial neural network (ANN) surrogate model, which was selected for its high accuracy in predicting structural responses, such as stress and total mass. This model was integrated with the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) to identify optimal designs balancing the maximum stress and total mass. The nine optimal crane designs were confirmed after conducting extra simulations to evaluate fatigue life, vertical deflection, and buckling safety factors. The most material-efficient design achieved a 28.7% weight reduction and a 28.6% improvement in fatigue life, and the design with the best mechanical performance showed the lowest Max stress (128.66 MPa) and the highest fatigue life (a 120.5% increase), with only a minor increase in mass. This range of optimized configurations offers adaptable, high-performance designs for industrial crane manufacturers. These results demonstrate that combining RSM, FEA, ANN, and NSGA-II is an effective tool for designing lightweight crane structures with high mechanical performance.