The efficiency in high-dimension optimization, premature convergence and noise robustness are common challenges that restrict the application of swarm-based algorithms in structural damage detection (SDD). In order to overcome these limitations, an improved grasshopper optimization algorithm (IGOA) based SDD approach assisted with adaptive region shrinkage strategies is proposed using modal strain energy assurance criterion. Firstly, the effectiveness of GOA in handling SDD is investigated, and an improved version is subsequently introduced with two adaptive region shrinkage strategies: adaptive Lévy flight mechanism and elite opposition-based learning, aiming to achieve more accurate and stable SDD results. Additionally, a novel objective function is defined, combining the frequency change ratio (FCR) and modal strain energy assurance criterion (MSEAC). To assess the performance of the proposed algorithm, a comparative analysis is conducted with four other selected swarm intelligent algorithms and GOAs with individual modification using four CEC benchmark functions. The results demonstrate that the proposed algorithm achieves improved convergence rate and accuracy. Furthermore, the effectiveness and efficiency of the proposed algorithm in SDD are validated through two numerical simulations and an experimental test. The SDD results obtained from both numerical simulations and experimental tests confirm the accuracy and robustness of the proposed method.