Engine Frame Model Update Based on A Levy Reverse Learning Dung Beetle Optimization Algorithm

This study introduces a novel approach for updating the dynamic properties of engine frame structures using an enhanced dung beetle optimization algorithm, referred to as the Levy Reverse Learning Dung Beetle Optimizer (LRDBO). The algorithm is strengthened by integrating a reverse learning method and a Levy flight technique, which improve both the initialization of the population and the way individual solutions are adjusted. These innovations lead to faster convergence and more precise optimization results. Additionally, a hybrid simulation platform combining MATLAB and Nastran is created to seamlessly link the optimization process with finite element analysis, offering a powerful toolset for structural design enhancement. When applied to update the engine frame model, the enhanced algorithm ensures that the error remains below 5%, in line with engineering specifications. The revised model accurately mirrors the dynamic characteristics of the actual structure, making it well-suited for real-world engineering tasks.