Adapted grasshopper optimization for damage detection of the structures using modal properties and the finite element model updating method

This paper presents a novel method to simulate the dynamic movement of grasshopper swarms, focusing on incorporating natural environmental factors—specifically wind and gravity—into the Grasshopper Optimization Algorithm (GOA). This enhanced version of the GOA algorithm, the Enhanced Grasshopper Optimization Algorithm (EGOA), is designed to improve both the exploration and exploitation phases of the optimization process. By integrating gravity and wind effects through specific transfer functions, the EGOA dynamically adjusts the swarm's behavior, enabling a more realistic and effective search mechanism within the algorithmic framework. The engineering significance of this work is highlighted through the application of EGOA to complex engineering problems, including damage identification in different structural systems. Since examples in damage detection of the structures are varied in size and degrees of freedom (DOF), choosing a capable optimization algorithm, which can handle this inverse problem without trapping in local optimums, is crucial. The objective function is defined based on frequencies and mode shapes of the structure in the damaged and undamaged states. Also, the design variables are the location and intensity of the damage in elements, which are achieved using optimization algorithms. Moreover, to simulate the real conditions in sensing the modal data, frequencies, and mode shapes are contaminated with noise. The effectiveness of the EGOA is rigorously tested against standard mathematical functions commonly used in optimization analyses. In this regard, five case studies were used for damage detection in various structures, including different 2D and 3D truss and frame structures with different scenarios of damage. The performance of EGOA is further validated through comparisons with established metaheuristic algorithms, including PSO, CSS, GWO, and CGO, demonstrating superior accuracy and stability in both mathematical benchmarks and structural damage detection. Additionally, experimental validation on a full-scale three-story steel frame confirms EGOA's practical efficacy, achieving precise damage localization and severity estimation with minimal false alarms and relative errors in modal properties. The results show that the proposed EGOA effectively addresses these complex nonlinear problems and maintains consistent performance across a large domain search space.